In r198030 the last in tree backend was converted to use MCInst for printing assembly. I removed support for the old printer in r198959. Out of tree targets have to lower MachineInstr to MCInst to use the new printer. Cheers, Rafael
Awesome. -eric On Fri, Jan 10, 2014 at 2:06 PM, Rafael Espíndola <rafael.espindola at gmail.com> wrote:> In r198030 the last in tree backend was converted to use MCInst for > printing assembly. I removed support for the old printer in r198959. > > Out of tree targets have to lower MachineInstr to MCInst to use the new printer. > > Cheers, > Rafael > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
And there was much rejoicing. -Jim On Jan 10, 2014, at 2:06 PM, Rafael Espíndola <rafael.espindola at gmail.com> wrote:> In r198030 the last in tree backend was converted to use MCInst for > printing assembly. I removed support for the old printer in r198959. > > Out of tree targets have to lower MachineInstr to MCInst to use the new printer. > > Cheers, > Rafael > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
On 10 January 2014 22:06, Rafael Espíndola <rafael.espindola at gmail.com>wrote:> In r198030 the last in tree backend was converted to use MCInst for > printing assembly. I removed support for the old printer in r198959. >And it only took what? Three years? Well done, everyone! --renato -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140110/7cebccc0/attachment.html>
On Jan 10, 2014, at 2:06 PM, Rafael Espíndola <rafael.espindola at gmail.com> wrote:> In r198030 the last in tree backend was converted to use MCInst for > printing assembly. I removed support for the old printer in r198959. > > Out of tree targets have to lower MachineInstr to MCInst to use the new printer.Awesome! Congrats Rafael, -Chris
>> Out of tree targets have to lower MachineInstr to MCInst to use the new printer. > > Awesome! Congrats Rafael,All credit to Venkatraman Govindaraju, I only noticed that the last one had been converted. Cheers, Rafael
Richard Pennington
2014-Jan-11 22:26 UTC
[LLVMdev] All backends now use the MC asm printer
On 01/10/2014 04:06 PM, Rafael Espíndola wrote:> In r198030 the last in tree backend was converted to use MCInst for > printing assembly. I removed support for the old printer in r198959. > > Out of tree targets have to lower MachineInstr to MCInst to use the new printer. > > Cheers, > Rafael >Are there any guidelines as to how the lowering should be done? I'd like to update the (defunct) Microblaze back end before I pull the latest TOT this week. -Rich
On Sat, Jan 11, 2014 at 5:26 PM, Richard Pennington <rich at pennware.com>wrote:> On 01/10/2014 04:06 PM, Rafael Espíndola wrote: > >> In r198030 the last in tree backend was converted to use MCInst for >> printing assembly. I removed support for the old printer in r198959. >> >> Out of tree targets have to lower MachineInstr to MCInst to use the new >> printer. >> >> Cheers, >> Rafael >> >> Are there any guidelines as to how the lowering should be done? I'd like > to update the (defunct) Microblaze back end before I pull the latest TOT > this week. >Venkatraman, I'm guessing you still have this fresh in your memory. Would you mind giving a hand? (bonus points for writing some documentation < http://llvm.org/docs/SphinxQuickstartTemplate.html> ;) -- Sean Silva> > -Rich > > _______________________________________________ > LLVM Developers mailing list > LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140112/ae125626/attachment.html>
> And it only took what? Three years? Well done, everyone!Well, there is still some work to be done before I would call the transition complete. A big part of the old logic is still present behind PrintAsmOperand. On targets that have an asm parser (and all of them should have one at some point), we should be parsing the inline asm string and then printing the operands as with regular instructions. Cheers, Rafael