On Thu, Dec 17, 2009 at 8:32 PM, Richard Henderson <r...@twiddle.net> wrote: >> funny how you seem to follow the same paths I did months ago :-) > > Heh. They are the obvious points for improvement in the emulation. > Hopefully we can get them finished this time, so that some third > person doesn't go through the same thing next year. =)
I agree :-) >> - cosmetics: don't use tabs and be sure not to have >> whitespaces at end of lines > > Grr, I thought I'd been extra careful about that. Fixed. > >> - the changes to tcg-op.h in patch 6 should go with patch 1 > > Done. > >> - outside of the small parameter handling, I'd prefer >> you provide a separate patch for tcg_out_brcond2; >> I don't think the changes using c1, c2 and c3 really >> belong to setcond/movcond. > > The i386 part has been split into 3 patches, though they > are sequentially dependent. > > I've left off the sparc backend patch, as well as the patches > for the translators. Let's take care of these first. Thanks, that will make things easier. > > r~ > --- > > Richard Henderson (6): This is a first pass of review. It's far complete (in particular no testing was done). Here is a summary of my comments: > tcg: Generic support for conditional set and conditional move. Needs cosmetics changes. > tcg: Add tcg_invert_cond. OK. > tcg-x86_64: Implement setcond and movcond. Some cosmetics and comments, but overall good. > tcg-i386: Implement small forward branches. I think this contains a bug. > tcg-i386: Simplify brcond2. I don't like the rewrite of brcond2. > tcg-i386: Implement setcond, movcond, setcond2. Not yet reviewed. Laurent > > elf.h | 2 + > tcg/README | 26 ++++- > tcg/i386/tcg-target.c | 278 ++++++++++++++++++++++++++++++++++++++-------- > tcg/tcg-op.h | 91 +++++++++++++++ > tcg/tcg-opc.h | 5 + > tcg/tcg.c | 23 +++- > tcg/tcg.h | 5 + > tcg/x86_64/tcg-target.c | 65 +++++++++++- > 8 files changed, 436 insertions(+), 59 deletions(-) > >
