* Added unification of existing conversions test as preparation for cl_khr_fp16 adaptation
* Unified initialization procedures for conversions test.
* Completed unification of data structures to handle cl_khr_fp16
* Added support for selective launch of the test
* Added half support for test_conversions, work in progres (issue #142, conversions)
* Added more work on halfs support for conversions test (issue #142, conversions)
* Added cosmetic corrections
* Added more cosmetic corrections before opening draft PR
* Added corrections related to pre-submit windows build
* Added more pre-build related corrections
* Added pre-submit ubuntu build related correction
* Added more pre-submit related corrections
* Divided structures into separate source files (issue #142, conversions)
* Added more corrections related to presubmit check
* Removed redeclarations due to presubmit check
* Added more corrections related to presubmit check arm build
* Added cosmetic correction
* Adapted modifications from related PR #1719 to avoid merging conflicts
* fixed clang format
* Added corrections related to code review (cl_khr_fp16 suuport according to issue #142)
* Corrections related to macos CI check fail
* fix for unclear clang format discrepancy
* More corrections related to code review (cl_khr_fp16 for conversions #142)
---------
Co-authored-by: Ewan Crawford <ewan@codeplay.com>
Avoid a shift by 64 on a `uint64_t`. The value resulting from the
spurious shift was overwritten later, so just avoid the shift in that
case.
Signed-off-by: Sven van Haastregt <sven.vanhaastregt@arm.com>
* Modernization of conversions test, preparation to handle cl_khr_fp16 extension
* Added missing virtual descructor
* Added corrections due to code review
* More separators removed
* Fixed clang format
* Added multiple corrections related to code review
* Corrected missing implicit test lost after modernization corrections
* Corrected single, selected test to limit number of unnecessary operations
* [NFC] Remove duplicate symbol declarations
Those are provided by testHarness.h.
Signed-off-by: Kévin Petit <kpet@free.fr>
* add missing include
---------
Signed-off-by: Kévin Petit <kpet@free.fr>
The host compiler will not always calculate reference values
the same, depending on optimization level. It generates
instructions that do not respond to CPU rounding mode in
the same way. Use QCOM rounding mode emulation to correctly
calculate reference values on aarch64.
