On 09/07/2016 11:48 PM, Benjamin Herrenschmidt wrote: > On Wed, 2016-09-07 at 14:13 +0200, Cédric Le Goater wrote: >> On 09/07/2016 01:08 PM, Benjamin Herrenschmidt wrote: >>> >>> On Wed, 2016-09-07 at 12:50 +0200, Cédric Le Goater wrote: >>>> >>>> This is a bit broader than Ben's patch which used >>>> PPC_SEGMENT_64B. >>>> it's basically !PPC_64B which includes the e5500. >>>> >>>> If so, here is a proposal below adding a new PPC_RFI in the >>>> "PowerPC Instructions types definitions" enum for that purpose. >>>> Not much bits left there. >>> >>> Why not stick to PPC_SEGMENT_64B ? >> >> I am trying to remove the rfi instruction from the set of the CPU >> and I think we need to introduce a new PPC_* bit for GEN_HANDLER to : > > What does it buy you instead of just having the test in the handler ?
not much. I will send the remaining bit of the original patch. Thanks, C. >> +GEN_HANDLER(rfi, 0x13, 0x12, 0x01, 0x03FF8001, PPC_RFI), >> >> we can also keep the test on PPC_SEGMENT_64B in the handler which >> works perfectly fine. >> >>> >>> rfi exists on all 32-bit processors and all non-Book3S (aka server >>> aka >>> segment/hash) 64-bit. So PPC_SEGMENT_64B is the test we want. >>> >>> IE. rfi does exist on e5500 >> >> ok. >> >> Cheers, >> C.
