Linux VP: Code merge + Single Core Variants #31
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
The top-levels of linux-vp(rv64) and linux32-vp(rv32) share a lot of common code. This redundancy makes changes that should apply to both vps error-prone. With this we merge these two toplevels to one. The distinction between linux-vp and linux32-vp is controlled by defines which are set in the corresponding CMakeLists. The change also includes a harmonization of the CombinedMemoryInterface for rv64 to match that of rv32 (mmu optional / passed by pointer) and according adjustments of tiny64 and tiny64-mc. -- [email protected] Institute for Complex Systems, JKU Linz https://ics.jku.at/
Newly added variants: * linux-sc-vp: RV64 with single worker core * linux32-sc-vp: RV32 with single worker core The motivation for this is mainly to speed up some experiments. For many experiments, a single core is sufficient and simulating a single core is much faster than simulating multiple cores. The new variants were introduced in the same way as it was done for the merge of linux-vp and linux32-vp: The distinction between is controlled by a define (NUM_CORES) which is set in the corresponding CMakeLists. It was also considered to make the number of cores configurable by a command-line parameter. However, because of the structure of the vp (templates) It was also considered to make the number of cores configurable via a command line parameter. However, the VP and SystemC make much use of template classes with the number of cores as a parameter. This could perhaps be worked around, but especially with sockets and bindings this quickly becomes messy and possibly inefficient. -- [email protected] Institute for Complex Systems, JKU Linz https://ics.jku.at/
Cirromulus
reviewed
Mar 7, 2023
| @@ -40,13 +40,15 @@ struct CombinedMemoryInterface : public sc_core::sc_module, | |||
| sc_core::sc_time dmi_access_delay = clock_cycle * 4; | |||
| std::vector<MemoryDMI> dmi_ranges; | |||
|
|
|||
| MMU &mmu; | |||
| MMU *mmu; | |||
There was a problem hiding this comment.
Why was this changed to a pointer instead of a reference? is there a case without an MMU?
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.
Hi,
with this I propose two changes which build up on on another:
The first one merges the top-level sources for linux-vp and linux32-vp to reduce code redundancy and the problems that come with it. (It can also be used as blue-print for similar cases)
The second one builds upon the first one and introduces single (worker) core variants of the linux-vp (linux-sc-vp (RV64) and linux32-sc-vp (RV32)), without adding much additional code.
Best regards,
Manfred
P.S.: Redarding usage, device-tree, etc: In the near future there will be a separate project that allows to easily create a basic setup for experiments with riscv-vp and Linux.