Portability fix-ups and generic chunk copy implementation. #51
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
added 2 commits
October 31, 2023 16:22
Generic paths for chunk ops and hash function. Add a cmake option to avoid hardware features so that the generic paths can be tested on x86.
Arguments which are uint64_t may be `%l` or `%ll` depending on target, which can produce warnings even if they're the same size. Using the PRI*64 macros makes this portable.
|
Thanks |
|
Hi @kornelski, This was intended to be the opener for a series of patches to improve RISC-V performance; but it got derailed and I've only recently been freed up to finish this off. The next patch, which actually includes RISC-V-specific code and a few hints at configuring the build appropriately, is here: After that I have more specialised repipelining of the decoder, but I need to do more work to either isolate it from the x86 path or to make the x86 path see the same benefits as I got under RISC-V. That's something I still need to figure out. |
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Outdated suggestions cannot be applied.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.
This adds a platform-independent path for chunkcopy.h for cases where the compiler is able to insert practical inline memcpy idioms similar to the ones hand-written for x86 and NEON. Depending on the architecture, this may depend on some
CFLAGSmagic to enable SIMD or unaligned data accesses.To test this on hardware which already has a dedicated chunk-copy implementation available, pass
-DBUILD_GENERIC_CODEand-DFORCE_CHUNK_COPYto cmake.