On Mon, Jun 22, 2020 at 5:31 PM Alexei Starovoitov <alexei.starovoitov@xxxxxxxxx> wrote: > > On Mon, Jun 22, 2020 at 05:09:04PM -0700, Andrii Nakryiko wrote: > > Extend original variable-length tests with a case to catch a common > > existing pattern of testing for < 0 for errors. Note because > > verifier also tracks upper bounds and we know it can not be greater > > than MAX_LEN here we can skip upper bound check. > > > > In ALU64 enabled compilation converting from long->int return types > > in probe helpers results in extra instruction pattern, <<= 32, s >>= 32. > > The trade-off is the non-ALU64 case works. If you really care about > > every extra insn (XDP case?) then you probably should be using original > > int type. > > > > In addition adding a sext insn to bpf might help the verifier in the > > general case to avoid these types of tricks. > > > > Signed-off-by: John Fastabend <john.fastabend@xxxxxxxxx> > > Please keep John's 'Author:' on the patch. > git commit --author= --amend > or keep 'From:' when you applied to your local git. I manually re-did the patch, because it wasn't applying cleanly. I'll use --author, though, I didn't know about it, thanks. > Also add your SOB after John's. > Even if you didn't change the patch at all. > Same thing if you've reworked the patch. ok, will do
