tianshilei1992 accepted this revision. tianshilei1992 added a comment. This revision is now accepted and ready to land.
In D113126#3212033 <https://reviews.llvm.org/D113126#3212033>, @jdoerfert wrote: > In D113126#3211935 <https://reviews.llvm.org/D113126#3211935>, > @tianshilei1992 wrote: > >> In D113126#3122947 <https://reviews.llvm.org/D113126#3122947>, @jdoerfert >> wrote: >> >>> In D113126#3122659 <https://reviews.llvm.org/D113126#3122659>, >>> @tianshilei1992 wrote: >>> >>>> I'm not convinced. `std::strlen` can do the job. Can you explain more why >>>> need it? >>> >>> Because the ident_t is on the device and you want to grab it from the host. >> >> So you mean `strlen` is not available on device? > > We have a device pointer to an ident_t object. We don't know the > corresponding host address and we are on the host. We can copy over the > ident_t with a single d2h but for the string we now need to either launch a > kernel that computes `strlen` on the device or copy it byte-by-byte, neither > are good options. Instead, we can simply embed the string length in the > otherwise unused ident_t field and copy it over with a second d2h memcpy. I see. Yeah, we don't have the map. Then that makes sense. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D113126/new/ https://reviews.llvm.org/D113126 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
