[llvm-commits] [PATCH, RFC] Medium code model support for 64-bit PowerPC
Bill Schmidt
wschmidt at linux.vnet.ibm.com
Tue Nov 27 09:20:00 PST 2012
On Tue, 2012-11-27 at 10:36 -0600, Will Schmidt wrote:
> On Mon, 2012-11-26 at 07:25 -0600, Bill Schmidt wrote:
> > Ping... Any further work needed before I commit this?
>
> I read through the latest patch, with your Uli inspired changes, and it
> looks good to me.
Thanks for the review!
>
> I'll note that test-wise, this is quite thorough, with coverage for
> loading (and storing) of the assorted variable types, as well as
> verifying the relocation tags for ; tentatively defined variable,
> external variable, function address, static variable scoped to function,
> file-scope static variable, constant pool.
>
> Cosmetically ... is the value 0x3F4FD4920B498CF0 used in the testcases
> special in any way?
Nope. Just due to a semi-random flailing of keys in the C source. The
value was initially specified as 9.71385298339204e-4, I believe, for no
rational reason whatsoever. :)
>
> Thanks,
> -Will
>
>
>
> > Thanks,
> > Bill
> >
> > On Fri, 2012-11-16 at 16:23 -0600, Bill Schmidt wrote:
> > > Here's a new revision of the patch. The changes are:
> > >
> > > (1) Add support for direct object file writing:
> > > - New relocation types in ELF.h.
> > > - New fixup kinds in PPCFixupKinds.h
> > > - Handle new fixups in PPCMCCodeEmitter.cpp, PPCAsmBackend.cpp, and
> > > PPCELFObjectWriter.cpp.
> > > - Add encoder method for LDrs in PPCInstr64Bit.td.
> > >
> > > (2) Small cleanups per private review from Roman Divacky:
> > > - Use GValue rather than GVal throughout, to more easily distinguish
> > > from GVar.
> > > - Use isa in preference to dyn_cast where appropriate in
> > > PPCISelDAGToDAG.cpp.
> > >
> > > I will be on vacation for the next week, but will occasionally monitor
> > > the list for any reviews. Thanks!
> > >
> > > Bill
> > >
> >
> > _______________________________________________
> > llvm-commits mailing list
> > llvm-commits at cs.uiuc.edu
> > http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> >
>
>
More information about the llvm-commits
mailing list