On 18 August 2015 at 22:29, Chen Gang <xili_gchen_5...@hotmail.com> wrote: > On 8/18/15 22:32, Peter Maydell wrote: >> On 18 August 2015 at 15:27, Chen Gang <xili_gchen_5...@hotmail.com> wrote: >>> Welcome any ideas, suggestions and completions. >> >> You should stop working on adding new features and instructions, >> and concentrate on getting a coherent set of patches for some >> subset of the instruction set reviewed and into QEMU. >> > > OK, thanks. It sounds good. But I guess, it is not executable: > > - I have already send a set of patches, but they are not integrated > into qemu (or not reviewed).
You need to concentrate on getting these in. That means: * check whether there are outstanding review comments (or trivial bugs you found yourself in the instructions covered by these parts) -- if so, then respin the patchset and resend it * 'ping' the patch series to remind people to review it (Specifically, IIRC, RTH needs to review the codegen bits of the integer patches.) > - I have to continue, although they are not integrated into (it means > I have to add new features and instructions, at present). If you do this then you are drawing the attention and time of reviewers away from the patches which are nearly ready to go into QEMU and towards the new stuff you post. This means that the older patchsets are less likely to move forward. > - For me, when tilegx qemu pass gcc testsuite, and finish floating > point insns in the preciser way, I guess, that is a reasonable point > to send new patches to qemu. This will result in a huge patchset which is very hard to review (and which is likely to get requests from me to split it up and send a smaller subset which is reviewable). thanks -- PMM
