Build JNI IFA switching helpers as C calls #2044
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -259,21 +259,17 @@ class RealRegisterManager | |
| TR::CodeGenerator* _cg; | ||
| }; | ||
|
|
||
| TR::Register * TR::S390CHelperLinkage::buildDirectDispatch(TR::Node * callNode, | ||
| TR::RegisterDependencyConditions **deps, | ||
| TR::Register *returnReg, | ||
| bool forceFastPath) | ||
| { | ||
| RealRegisterManager RealRegisters(cg()); | ||
| bool isHelperCallWithinICF = deps != NULL; | ||
|
There was a problem hiding this comment. Could we leave this TODO around? This is definitely something we still want to do once we refactor the runtime helpers into a sane table with properties. There was a problem hiding this comment. TODO is added back |
||
| // TODO: Currently only jitInstanceOf and IFA helper calls are fast path helpers. Need to modify following condition if we add support for other fast path only helpers | ||
| // Having a fast path helper call also implies that there is no environment pointer setup in GPR5 | ||
| bool isFastPathOnly = callNode->getOpCodeValue() == TR::instanceof || forceFastPath; | ||
|
|
||
| traceMsg(comp(),"%s: Internal Control Flow in OOL : %s\n",callNode->getOpCode().getName(),isHelperCallWithinICF ? "true" : "false" ); | ||
| for (int i = TR::RealRegister::FirstGPR; i <= TR::RealRegister::LastHPR; i++) | ||
| { | ||
|
|
@@ -344,6 +340,7 @@ TR::Register * TR::S390CHelperLinkage::buildDirectDispatch(TR::Node * callNode, | |
| // Storing Java Stack Pointer | ||
| javaStackPointerRegister = cg()->getStackPointerRealRegister(); | ||
| cursor = generateRXInstruction(cg(), TR::InstOpCode::getStoreOpCode(), callNode, javaStackPointerRegister, generateS390MemoryReference(vmThreadRegister, offsetJ9SP, cg())); | ||
|
|
||
| #if defined(J9ZOS390) | ||
| padding += 2; | ||
| // Loading DSAPointer Register | ||
|
|
@@ -364,6 +361,21 @@ TR::Register * TR::S390CHelperLinkage::buildDirectDispatch(TR::Node * callNode, | |
| } | ||
| TR::SymbolReference * callSymRef = callNode->getSymbolReference(); | ||
| void * destAddr = callNode->getSymbolReference()->getSymbol()->castToMethodSymbol()->getMethodAddress(); | ||
|
|
||
| #if defined(J9ZOS390) | ||
|
There was a problem hiding this comment. @NigelYiboYu I think my previous comment was for this section. For now we need this part for calling out helpers written in C that means we need to load environment for instanceOf as well. Is that covered? Looking through your other changes. There was a problem hiding this comment. @fjeremic I am thinking of moving some of the context switching code from helper Glue code to in body and see the effect. I expect a small effect on the method footprint. If the effect is not much then we can move this inlined. I believe x86 already do this. We can make our helper linkage more generic this way. There was a problem hiding this comment. @r30shah let's give it a try and see if it has any negative effects on performance. There was a problem hiding this comment.
There was a problem hiding this comment. Actually good point @r30shah. Thinking about this again we have duplicate information here. The helper index is already stored in This information is already available through the There was a problem hiding this comment. The extra helper index is removed. There was a problem hiding this comment. Update: Environment pointer is now only loaded if the C helper is fast-path. Non-fastPath helpers jump to glue code, which does its own environment handling. |
||
| // Fast path helper calls are real C function calls, whereas non-fast-path calls jump to picBuilder glue code | ||
| // glue code does its own environment handling and does not need a environment loading here. | ||
| if(isFastPathOnly) | ||
| { | ||
| TR_RuntimeHelper helperIndex = static_cast<TR_RuntimeHelper>(callNode->getSymbolReference()->getReferenceNumber()); | ||
| TR_ASSERT_FATAL(helperIndex < TR_S390numRuntimeHelpers, "Invalid helper index in c helper node\n"); | ||
|
|
||
| // XPLINK GPR5 is an environment register. | ||
| intptrj_t environment = reinterpret_cast<intptrj_t>(TOC_UNWRAP_ENV(runtimeHelpers.getFunctionPointer(helperIndex))); | ||
| genLoadAddressConstant(cg(), callNode, environment, javaStackPointerRegister); | ||
| } | ||
| #endif | ||
|
|
||
| cursor = new (cg()->trHeapMemory()) TR::S390RILInstruction(TR::InstOpCode::BRASL, callNode, regRA, destAddr, callSymRef, cg()); | ||
| cursor->setDependencyConditions(preDeps); | ||
| if (isFastPathOnly) | ||
|
|
||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think we'll need to modify the CMake builds as well.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't see CMake lists contain these
cppfile names. It seems to operate on directories for z codegen.