search for: indermediate

Hi! The following thoughts about the llvm architecture I'd like to share with you (from the perspective of a user): - If a backend has no vector support, then I wonder why there is no de-vectorization pass that operates on indermediate llvm-ir. I would think that if you use such a target then you have to insert a target independent pass before it that it does not have to care about vector code. The advantage is that constant vector components can already be handled by instcombine. what do you think? - If the integer width of...

...following thoughts about the llvm architecture I'd like to share >>> with you >>> (from the perspective of a user): >>> >>> - If a backend has no vector support, then I wonder why there is no >>> de-vectorization >>> pass that operates on indermediate llvm-ir. I would think that if you >>> use such a target >>> then you have to insert a target independent pass before it that it does >>> not have to >>> care about vector code. The advantage is that constant vector components >>> can already >>...

...i! >> >> The following thoughts about the llvm architecture I'd like to share >> with you >> (from the perspective of a user): >> >> - If a backend has no vector support, then I wonder why there is no >> de-vectorization >> pass that operates on indermediate llvm-ir. I would think that if you >> use such a target >> then you have to insert a target independent pass before it that it does >> not have to >> care about vector code. The advantage is that constant vector components >> can already >> be handled by instc...

...Jochen Wilhelmy wrote: > Hi! > > The following thoughts about the llvm architecture I'd like to share > with you > (from the perspective of a user): > > - If a backend has no vector support, then I wonder why there is no > de-vectorization > pass that operates on indermediate llvm-ir. I would think that if you > use such a target > then you have to insert a target independent pass before it that it does > not have to > care about vector code. The advantage is that constant vector components > can already > be handled by instcombine. what do you thin...

...t the llvm architecture I'd like to share >>>> with you >>>> (from the perspective of a user): >>>> >>>> - If a backend has no vector support, then I wonder why there is no >>>> de-vectorization >>>> pass that operates on indermediate llvm-ir. I would think that if you >>>> use such a target >>>> then you have to insert a target independent pass before it that it does >>>> not have to >>>> care about vector code. The advantage is that constant vector components >>>>...

...I'd like to share >>>>> with you >>>>> (from the perspective of a user): >>>>> >>>>> - If a backend has no vector support, then I wonder why there is no >>>>> de-vectorization >>>>> pass that operates on indermediate llvm-ir. I would think that if you >>>>> use such a target >>>>> then you have to insert a target independent pass before it that it does >>>>> not have to >>>>> care about vector code. The advantage is that constant vector components &...