Hi Dmitry, thanks for the quick reply :) I've already filed an issue which covers this while including an additional quesion I had. See #1438. If you would like an additional issue more targeted at the CFG (I guess ControlFlowGraph?) issue I'm happy to create another one. Additionally, if you think this is something a person new to compilers can tackle I'm also happy to have a look into this, given some pointers on where to start looking.
Cheers, Christoph On Friday, March 15, 2019 at 5:53:19 PM UTC+1, Dmitry Babokin wrote: > > No, we don't have mechanisms like __builtin_unreachable(). I'm not sure if > extracting some data flow facts from "if ( none ( S > 25 ) ) > flag_unreachable;" to apply to "if (S > 25)" is good idea, but we should > definitely think about mechanisms to fine tune CFG. > > Would be good if you file an issue with such request. > > Dmitry. > > On Thu, Mar 14, 2019 at 7:40 AM Christoph Hasse <[email protected] > <javascript:>> wrote: > >> Hi All, >> >> I'm new to ISPC and have been playing around with it a bit to evaluate if >> it's a possible solution for our codebase. So far I'm liking it ;) >> >> A general question I've not been able to find a solution too is if there >> is something similar to __builtin_unreachable() in GCC. >> >> Looking at this simple example https://ispc.godbolt.org/z/Meee-G I'd >> like to disable the code path where (S>25) is false for all instances in a >> gang to get rid of the additional check in the assembly. >> Keep in mind that this is a simple example so it's not so much about if >> this will really increase performance, but more about if I can do it. >> >> I was hoping for something like: if ( none ( S > 25 ) ) flag_unreachable; >> >> Cheers, >> Christoph >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Intel SPMD Program Compiler Users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> For more options, visit https://groups.google.com/d/optout. >> > -- You received this message because you are subscribed to the Google Groups "Intel SPMD Program Compiler Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
