search for: passmanagerintern

Hello, (fairly new around here, apologies if the question has an obvious answer) I've been studying the structure of the new PassManager in the hopes of using it soon. After watching a couple of talks (Chandler / Sergeev), I decided to start with the Concept/Model classes in PassManagerInternal.h While I could make sense of almost everything there, two details caught my attention: 1- The PassConcept abstract class has a function `run = 0;` which returns an object of type `PreservedAnalysis`. Such object is forward-declared in this header, but it is defined in PassManager.h This is th...

...thers' thoughts are on this issue. -- Sean Silva On Sun, Jul 24, 2016 at 2:58 AM, Sean Silva <chisophugis at gmail.com> wrote: > I've started looking specifically at the existing code and how it needs to > be changed. It seems like the concept-based polymorphism stuff in > PassManagerInternal.h actually don't need to be changed that much. > AnalysisPassConcept and AnalysisResultConcept need to be changed to take a > type-erased IRUnit (e.g. a void* or something). AnalysisPassModel and > AnalysisResultModel (which inherit from their respective abstract base > classes I...

...sue. > > -- Sean Silva > > On Sun, Jul 24, 2016 at 2:58 AM, Sean Silva <chisophugis at gmail.com> wrote: > >> I've started looking specifically at the existing code and how it needs >> to be changed. It seems like the concept-based polymorphism stuff in >> PassManagerInternal.h actually don't need to be changed that much. >> AnalysisPassConcept and AnalysisResultConcept need to be changed to take a >> type-erased IRUnit (e.g. a void* or something). AnalysisPassModel and >> AnalysisResultModel (which inherit from their respective abstract base &gt...

...> >> On Sun, Jul 24, 2016 at 2:58 AM, Sean Silva <chisophugis at gmail.com> >> wrote: >> >>> I've started looking specifically at the existing code and how it needs >>> to be changed. It seems like the concept-based polymorphism stuff in >>> PassManagerInternal.h actually don't need to be changed that much. >>> AnalysisPassConcept and AnalysisResultConcept need to be changed to take a >>> type-erased IRUnit (e.g. a void* or something). AnalysisPassModel and >>> AnalysisResultModel (which inherit from their respective abstr...

The more closely I look at this, the more it seems like there may be a useful incremental step in the transition to the new PM: use the new PM analysis machinery in the old PM. If this is possible, it will simplify the old PM and (hopefully) allow an incremental transition to the new PM instead of a flag day transition for the switch. I.e., AFAICT, the new PM transition is essentially about 2