[spi_device] Add support for 1r1w RAMs and parity init #20942
Merged
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
Signed-off-by: Robert Schilling <[email protected]>
Signed-off-by: Michael Schaffner <[email protected]>
Signed-off-by: Michael Schaffner <[email protected]>
Signed-off-by: Michael Schaffner <[email protected]>
Some core files were missing, and there were also some syntax errors. Signed-off-by: Alexander Williams <[email protected]>
a-will
added
Component:DV
a-will
requested review from
jdonjdon and
HU90m
and removed request for
a team
hw/ip/spi_device/rtl/spid_dpram.sv
Outdated
| spi2sys_rd_req, | ||
| spi2sys_rd_addr | ||
| }; | ||
| end else if (SramType == SramType1r1w) begin : ram1r1w |
There was a problem hiding this comment.
nitnit: we should fix up these block labels as the linter says
msfschaffner
approved these changes
Jan 23, 2024
Yes, I squashed that all down, since it was noise. |
Add 1r1w RAM configuration as an option for spi_device for tech nodes where the 2p RAM configuration is not available. Make the 2p RAM have the same access controls as the 1r1w RAM, so the two behave the same way. Also add word initialization circuitry on the SPI side, to init parity. The SPI -> core buffer for the payload uses parity and SW has no way of initializing it since the the write port is in the SPI domain. Since the SPI side writes the payload byte by byte, we need to guard against partially initialized 32bit wordd, because these could cause TL-UL bus errors upon readout. Unfortunately, an initialization circuit that initializes the entire SRAM on the SPI clock domain is infeasible since that clock is only intermittently available. Hence, we keep track of uninitialized words using a valid bit array, and upon the first write to a word, uninitialized bytes are set to zero if the write operation is a sub-word write op. Note that in this commit, DV tests have focused much more on the 2p variant. Signed-off-by: Alexander Williams <[email protected]> Co-authored-by: Michael Schaffner <[email protected]>
Add params for the DPRAM offsets so they are available in C headers for software. Signed-off-by: Alexander Williams <[email protected]>
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.
Add 1r1w RAM configuration as an option for spi_device for tech nodes where the 2p RAM configuration is not available. Make the 2p RAM have the same access controls as the 1r1w RAM, so the two behave the same way.
Also add word initialization circuitry on the SPI side, to init parity. The SPI -> core buffer for the payload uses parity and SW has no way of initializing it since the the write port is in the SPI domain. Since the SPI side writes the payload byte by byte, we need to guard against partially initialized 32bit wordd, because these could cause TL-UL bus errors upon readout. Unfortunately, an initialization circuit that initializes the entire SRAM on the SPI clock domain is infeasible since that clock is only intermittently available. Hence, we keep track of uninitialized words using a valid bit array, and upon the first write to a word, uninitialized bytes are set to zero if the write operation is a sub-word write op.
Note that in this PR, DV tests have focused much more on the 2p variant, though.
This is largely a set of cherry-picks from integrated_dev, after resolving conflicts. However, the last commit adds fixed "parameters" for DPRAM offsets and makes them available as macros to C.
Resolves #5369