Alexei Starovoitov <alexei.starovoitov@xxxxxxxxx> wrote: > On Thu, Oct 27, 2022 at 7:46 AM <dthaler1968@xxxxxxxxxxxxxx> wrote: > > > > + > > +Thus the 64-bit immediate value is constructed as follows: > > + > > + imm64 = imm + (next_imm << 32) > > Are you sure this is correct considering that 'imm' > was defined earlier in the doc as sign extended? > Maybe use: > imm64 = (u32)imm | ((u64)(u32)next_imm) << 32) ? I agree that would be clearer, will update. Dave
