-
Notifications
You must be signed in to change notification settings - Fork 4.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Enable AVX512 Additional 16 SIMD Registers #79544
Commits on Jan 17, 2023
-
Change regMask_enum and regMaskTP to unsigned __int64_t on AMD64.
This allows for more registers to be encoded in the register allocator.
Configuration menu - View commit details
-
Copy full SHA for 9da5b6d - Browse repository at this point
Copy the full SHA 9da5b6dView commit details -
Add upper 16 SIMD registers to allocator.
Commit includes refactoring code to use `const instrDesc *` instead of `instruction` so information about when EVEX is needed (due to high SIMD registers) is available to the emitter.
Configuration menu - View commit details
-
Copy full SHA for 6c6c884 - Browse repository at this point
Copy the full SHA 6c6c884View commit details -
Configuration menu - View commit details
-
Copy full SHA for bd6d2a5 - Browse repository at this point
Copy the full SHA bd6d2a5View commit details -
Configuration menu - View commit details
-
Copy full SHA for 47ff9ca - Browse repository at this point
Copy the full SHA 47ff9caView commit details -
Limit high SIMD reg to compatible intrinsics and gentree nodes.
Commit constrains certain hw intrinsics and gentree nodes to use lower SIMD registers even if upper SIMD registers are available due to limitations of EVEX encoding for certain instructions. For example, SSE `Reciprocal` lowers to `rcpps` which does not have an EVEX encoding form, hence, we cannot allow that hw intrincis node to use a high SIMD register. These intrinsics are marked with `HW_Flag_NoEvexSemantics`. Other such intructions related to masking (typically marked with `HW_Flag_ReturnsPerElementMask`) also have similar issues (though they can be replaced with the EVEX k registers and associated masking when implemented). In addition, the callee/calleer save registers have also been adjusted to properly handle the presence and absence of AVX512 upper simd registers at runtime.
Configuration menu - View commit details
-
Copy full SHA for 9cabef6 - Browse repository at this point
Copy the full SHA 9cabef6View commit details -
Configuration menu - View commit details
-
Copy full SHA for 9f711ba - Browse repository at this point
Copy the full SHA 9f711baView commit details -
Configuration menu - View commit details
-
Copy full SHA for f29c146 - Browse repository at this point
Copy the full SHA f29c146View commit details -
Configuration menu - View commit details
-
Copy full SHA for b95d296 - Browse repository at this point
Copy the full SHA b95d296View commit details -
Configuration menu - View commit details
-
Copy full SHA for 6a4dcd1 - Browse repository at this point
Copy the full SHA 6a4dcd1View commit details -
Configuration menu - View commit details
-
Copy full SHA for d1f2fdb - Browse repository at this point
Copy the full SHA d1f2fdbView commit details -
Update src/coreclr/jit/compiler.h
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for c7807c7 - Browse repository at this point
Copy the full SHA c7807c7View commit details -
Update src/coreclr/jit/compiler.cpp
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 0b6cb55 - Browse repository at this point
Copy the full SHA 0b6cb55View commit details -
Update src/coreclr/jit/gentree.cpp
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for a45163c - Browse repository at this point
Copy the full SHA a45163cView commit details -
Update src/coreclr/jit/hwintrinsic.h
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 730b4eb - Browse repository at this point
Copy the full SHA 730b4ebView commit details -
Update src/coreclr/jit/target.h
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 34d6199 - Browse repository at this point
Copy the full SHA 34d6199View commit details -
Update src/coreclr/jit/emitxarch.cpp
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 395f5e4 - Browse repository at this point
Copy the full SHA 395f5e4View commit details -
Configuration menu - View commit details
-
Copy full SHA for 852e051 - Browse repository at this point
Copy the full SHA 852e051View commit details -
Configuration menu - View commit details
-
Copy full SHA for 7238b49 - Browse repository at this point
Copy the full SHA 7238b49View commit details -
Configuration menu - View commit details
-
Copy full SHA for 2b56df7 - Browse repository at this point
Copy the full SHA 2b56df7View commit details -
Configuration menu - View commit details
-
Copy full SHA for fa1a550 - Browse repository at this point
Copy the full SHA fa1a550View commit details -
Configuration menu - View commit details
-
Copy full SHA for 564dc81 - Browse repository at this point
Copy the full SHA 564dc81View commit details -
Configuration menu - View commit details
-
Copy full SHA for 5ea489d - Browse repository at this point
Copy the full SHA 5ea489dView commit details
Commits on Jan 18, 2023
-
Configuration menu - View commit details
-
Copy full SHA for e60aece - Browse repository at this point
Copy the full SHA e60aeceView commit details
Commits on Jan 26, 2023
-
Update src/coreclr/jit/compiler.h
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for b28a231 - Browse repository at this point
Copy the full SHA b28a231View commit details -
Update src/coreclr/jit/lsra.cpp
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 37af7c3 - Browse repository at this point
Copy the full SHA 37af7c3View commit details -
Update src/coreclr/jit/compiler.h
Co-authored-by: Bruce Forstall <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for e9be8f8 - Browse repository at this point
Copy the full SHA e9be8f8View commit details -
Configuration menu - View commit details
-
Copy full SHA for f680448 - Browse repository at this point
Copy the full SHA f680448View commit details -
Merge branch 'avx512-upper-regs' of https://github.com/anthonycanino/…
…runtime into avx512-upper-regs
Configuration menu - View commit details
-
Copy full SHA for 7416501 - Browse repository at this point
Copy the full SHA 7416501View commit details
Commits on Jan 31, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 73f43b6 - Browse repository at this point
Copy the full SHA 73f43b6View commit details
Commits on Feb 2, 2023
-
Configuration menu - View commit details
-
Copy full SHA for c6f1a90 - Browse repository at this point
Copy the full SHA c6f1a90View commit details
Commits on Feb 3, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 91cf3db - Browse repository at this point
Copy the full SHA 91cf3dbView commit details -
Configuration menu - View commit details
-
Copy full SHA for 228c0c5 - Browse repository at this point
Copy the full SHA 228c0c5View commit details
Commits on Feb 6, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 5490617 - Browse repository at this point
Copy the full SHA 5490617View commit details
Commits on Feb 8, 2023
-
Use inline accessor functions instead of macros
Convert from macros to accessor functions for RBM_ALLFLOAT, RBM_FLT_CALLEE_TRASH, CNT_CALLEE_TRASH_FLOAT. Convert LSRA use of ACTUAL_REG_COUNT to AVAILABLE_REG_COUNT, and create an accessor for that value for AMD64 as well.
Configuration menu - View commit details
-
Copy full SHA for 3c7acdb - Browse repository at this point
Copy the full SHA 3c7acdbView commit details -
Merge pull request #6 from BruceForstall/anthonycanino_avx512-upper-r…
…egs-with-reg-accessors Use inline accessor functions instead of macros
Configuration menu - View commit details
-
Copy full SHA for e42db99 - Browse repository at this point
Copy the full SHA e42db99View commit details -
Configuration menu - View commit details
-
Copy full SHA for 10b4703 - Browse repository at this point
Copy the full SHA 10b4703View commit details