<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Hi Jeffrey,<div><br><div><div>On 2009-06-22, at 01:49, Jeffrey Yasskin wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div>I need to add thread-switching support to Unladen Swallow's JIT, and LLVM's safe point support looks like a good way to get code into all the right places. However, <a href="http://llvm.org/docs/GarbageCollection.html#collector-algos">http://llvm.org/docs/GarbageCollection.html#collector-algos</a> points out that there's no way to emit code at safe points yet, and there are no loop safe points at all. So I'll be trying to implement them.<br><br>Is there anything I should know before starting?<br></div></blockquote><div><br></div><div>Sounds like you've got the right idea.</div><div><br></div><blockquote type="cite"><div>One way to do this might be to add a FunctionPass to LLVMTargetMachine::addCommonCodeGenPasses() alongside createGCLoweringPass(), which would insert user-defined code for safe points. Unfortunately, code inserted there would stick around in the IR after the machine code was emitted, and if the function were JITted again, we'd get duplicate safe points.<br></div></blockquote><div><br></div><div>Unfortunately, I don't believe this is workable. It would make this work much easier if it were.</div><div><br></div><blockquote type="cite"><div>Another way to do it might be to add a MachineFunction pass next to createGCMachineCodeAnalysisPass() (or instead of it), which could emit appropriate MachineInstructions to implement the safe point. This, of course, forces safe points to be written in MachineInstructions instead of IR instructions, which isn't ideal.<br></div></blockquote><div><br></div><div>I think this is the way to go, though it's also the most involved. My primary rationale is that code generation can hack on the CFG, even introducing loops where there were none expressed in the IR. It could be that I'm being unnecessarily pessimistic on this point, though.</div><div><br></div><div>As a specific example of the code generator hacking on the CFG, take atomic operations which expand to loops on architectures which use load-reserved/store-conditional to implement these primitives. It may not be necessary or desirable to add safe points to these loops, but perhaps should be preferred on the basis o correctness.</div><div><br></div><div>As another example, consider a 64-bit integer divide on a 32-bit architecture expanding to a libcall. Some, but perhaps not all, collection algorithms would want to emit safe point code for this call, but it simply does not exist in the IR to instrument.</div><div><br></div><div>Also, code injection of the form 'give me 8 bytes of noops at each safe point' and 'insert a cold instruction sequence at the end of the function' are best expressed at the machine code level. Safe points are hot code and unusual, target-specific techniques are regularly used with them if you survey the literature, so a design which accommodates that reality is preferred, even though hacking on the MachineFunction representation is less pleasant than the IR.</div><div><br></div><div>One element of this design that is desirable from a design perspective is that it preserves the original IR. Chris has said that it's a long-term goal of LLVM to not mangle the Function during code generation, and this moves in that direction instead of regressing.</div><div><br></div><blockquote type="cite"><div>Another way might be to run a pass over the IR inserting llvm.safepoint() calls, which could be implemented as a function in the module. Then we'd want a MachineFunction pass to inline this for us during codegen. The llvm.safepoint() calls could be easily identified and removed if the IR needs to be re-used.</div></blockquote><br></div><div>I see this as fairly equivalent to the first option.</div><div><br></div><div>Also, regardless, stop point markers (a label is actually generated) need to survive as such into the MachineFunction representation else we'll never be able to generate a register map.</div><div><br></div><div>Hope that helps,</div><div>Gordon</div><div><br></div><div><br></div><div><div>P.S. There's an interesting circularity which I may not have accounted for in the original design: If code is injected at each safe point, and a call instruction is injected, do we need to generate <i>another</i> safe point for that call? Clearly, the expansion of a safe point cannot be recursive with itself; but I think that we should allow generating a register map at the return address of that call, just as some collectors may want to instrument the libcall case discussed above.</div><div><br></div><div>Actually, this distinction between safe points for inserting code and safe points for frame maps is probably is a critical design issue for your use case. Our current definition of a safe point is at the return address of a call instruction, which is precisely what's required to call the stack. This is NOT the location where you want to add a call to your runtime.</div></div></div></body></html>