[SYCL] Added support of rounding modes for non-host devices#1463
Conversation
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
|
Tagging @Ruyk to take a look at changes at |
sycl/include/CL/sycl/types.hpp
Outdated
| template <typename T, typename R, rounding_mode roundingMode> | ||
| detail::enable_if_t<!std::is_same<T, R>::value && | ||
| (is_int_to_int<T, R>::value && | ||
| !(is_int_to_from_uint<T, R>::value) || |
There was a problem hiding this comment.
This line seem cause build error, please double-check
There was a problem hiding this comment.
I forgot to delete it.
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
|
This connects to @Naghasan 's work |
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
clang/lib/Sema/SPIRVBuiltins.td
Outdated
| } | ||
|
|
||
| foreach IType = [Char, Short, Int, Long] in { | ||
| foreach IType = [TrueChar, Short, Int, Long] in { |
There was a problem hiding this comment.
changing https://github.com/intel/llvm/pull/1463/files#diff-35388f048c1dc54db0a9491da9317c69R290 to char is the actual fix I believe. This change will trigger unwanted implicit conversions.
Char is bound to std::int8_t (always signed char) while TrueChar is the actual char type which can be unsigned. I added TrueChar as a hack for the __spirv_ocl_printf but I don't think it should be kept.
There was a problem hiding this comment.
Sorry the link don't seem to work properly, this line
llvm/clang/lib/Sema/SPIRVBuiltins.td
Line 290 in e875fbf
schar to char so the function name can be built correctly (with e875fbf it is now post-fixing with _Rschar)
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
| half8{+2.3f, +2.5f, +2.7f, -2.3f, -2.5f, -2.7f, 0.f, 0.f}); | ||
|
|
||
| return 0; | ||
| } | ||
| } No newline at end of file |
There was a problem hiding this comment.
Don't see much value in these changes.
Suggest keeping original code.
There was a problem hiding this comment.
Only part of this comment was resolved.
There was a problem hiding this comment.
I don't understand, why they are different?
There was a problem hiding this comment.
You seem to delete line ending character.
Signed-off-by: Aleksander Fadeev <aleksander.fadeev@intel.com>
| // Unsupported: cuda | ||
| // The test fails sporadically on cuda. See https://github.com/intel/llvm/issues/1508 for more details. | ||
|
|
There was a problem hiding this comment.
Please, revert changes in this file.
d9b2d52 already disabled the test.
There was a problem hiding this comment.
Do I need to do a rebase?
There was a problem hiding this comment.
No. CI system does the merge with tip of the sycl branch before testing.
bader
left a comment
There was a problem hiding this comment.
LGTM.
The description doesn't say this, but this patch implements conversion from FP to Int only. What about Int to FP?
I implemented this in the other PR. |
|
@Naghasan, @erichkeane, @Fznamznon, @turinevgeny, please approve if you don't have any other comments. |
|
LGTM. |
|
@turinevgeny, you must approve using GitHub process: https://help.github.com/en/github/collaborating-with-issues-and-pull-requests/approving-a-pull-request-with-required-reviews. |
…versioning * origin/sycl: [XPTI][Framework] Reference implementation of the Xpti framework to be used with instrumentation in SYCL (intel#1557) [SYCL] Initial ABI checks implementation (intel#1528) [SYCL] Support connection with multiple plugins (intel#1490) [SYCL] Add a new header file with the reduction class definition (intel#1558) [SYCL] Add test for SYCL kernels with accessor and spec constant (intel#1536) [SYCL][CUDA] Move interop tests (intel#1570) [Driver][SYCL] Remove COFF object format designator for Windows device compiles (intel#1574) [SYCL] Fix conflicting visibility attributes (intel#1571) [SYCL][DOC] Update the SYCL Runtime Interface document with design details (intel#680) [SYCL] Improve image accessors support on a host device (intel#1502) [SYCL] Make queue's non-USM event ownership temporary (intel#1561) [SYCL] Added support of rounding modes for non-host devices (intel#1463) [SYCL] SemaSYCL significant refactoring (intel#1517) [SYCL] Support 0-dim accessor in handler::copy(accessor, accessor) (intel#1551)
Implementing rounding models for cl::sycl::vec type for non host devices.
Signed-off-by: Aleksander Fadeev aleksander.fadeev@intel.com