WIP: Adding Sycl Support to CYK Calculations #240
Conversation
src/cpp.cc
Outdated
| #include "cpp.hh" | ||
|
|
||
| #include <cmath> | ||
| #include <list> | ||
| #include <set> | ||
| #include <string> | ||
| #include <tuple> | ||
| #include <utility> | ||
| #include <vector> | ||
| #include <string> | ||
| #include <list> | ||
| #include <boost/tokenizer.hpp> | ||
|
|
||
| #include "cpp.hh" | ||
|
|
||
| #include "ast.hh" | ||
| #include "cyk.hh" | ||
| #include "expr.hh" | ||
| #include "expr/new.hh" | ||
| #include "fn_def.hh" | ||
| #include "grammar.hh" | ||
| #include "options.hh" | ||
| #include "outside/codegen.hh" | ||
| #include "statement.hh" | ||
| #include "statement/backtrace_decl.hh" | ||
| #include "statement/fn_call.hh" | ||
| #include "statement/hash_decl.hh" |
There was a problem hiding this comment.
This is a bit confusing. Why did you reorder the imports? Current tests break because of some import issues, is this here the cause? Hard to debug. Could you please revert as many lines as possible to show what really changed?
src/cpp.cc
Outdated
| lines.begin(); i != lines.end(); ++i) { | ||
| lines.begin(); | ||
| i != lines.end(); ++i) { |
There was a problem hiding this comment.
why this change?
If you care about code style - and I am a fan of this - we might want to split your PR into one that adds functionality and a second one only addressing code style changes
|
|
||
| #ifndef SRC_CPP_HH_ | ||
| #define SRC_CPP_HH_ | ||
|
|
||
| #include <vector> | ||
| #include <string> | ||
| #include <list> | ||
| #include <string> | ||
| #include <vector> | ||
|
|
||
| #include "operator.hh" | ||
|
|
||
| #include "printer.hh" | ||
| #include "codegen.hh" | ||
|
|
||
| #include "table.hh" | ||
| #include "symbol_fwd.hh" | ||
| #include "operator.hh" | ||
| #include "para_decl_fwd.hh" | ||
| #include "printer.hh" | ||
| #include "product.hh" | ||
| #include "symbol_fwd.hh" | ||
| #include "table.hh" |
There was a problem hiding this comment.
same critique applies here as for src/cpp.cc. This PR is overly hard to understand as the many thousand lines of code that changed just because of code-style are very hard to differentiate from conceptional changes :-/
| new Type::String, "test_value"); | ||
|
|
||
| fn_cyk->stmts.push_back( | ||
| new Statement::SYCL_Buffer_Decl(new Type::Int, dimension, value, value)); |
There was a problem hiding this comment.
if you are going to use such a statement, don't you have to include according cycl/cycl.hpp header files?! The currently failing tests point to exactly these headers being absent.
There was a problem hiding this comment.
We furthermore might want to think about a user switch/flag to decide if generated code shall contain sycl "extensions" or not?!
There was a problem hiding this comment.
yes, this is the plan for this week, implementing headers and the proper loop
2 participants
This is the pull request that should be used to discuss the sycl specific changes. It is currently only a draft.