Re: [Qemu-devel] [PATCH 1/6 v7] target-tilegx: Firstly add TILE-Gx with minimized features

[Peter Maydell <peter.maydell@linaro.org>]

On 20 March 2015 at 23:57, Chen Gang <xili_gchen_5257@hotmail.com> wrote:
> For 'opcode_tilegx.h', it comes from Linux kernel which is already more
> than 1000 lines. For me, I still suggest to let it within one patch.

Right, for that sort of special case you should put it in
a single patch which is just "Import copy of whatever.h from Linux
kernel", and which has nothing else in it except for the imported
file or files, and which makes no QEMU specific changes to those
files. Then people can skim through that patch quickly knowing
they don't have to look for any changes or extras that do need
review. (If there is some tweak that needs to be made for QEMU's
purposes that can go in a subsequent patch.)

> For 'translate.c', I try to split it into 3 patches:
>
>  - parsing bundle and decode to the first level of pipe,
>
>  - parsing 2nd level of pipe.
>
>  - generate tcg code.
>
> For the left files:
>
>  - 'configure', 'Makefile.obj', 'tilegx-linux-user.mak' are in a patch.
>
>  - 'cpu*' are in a patch
>
>  - 'help.*' are in a patch.

I would suggest that rather than having "cpu* in one patch
and help* in another" (ie a split by file), you split things
functionally, so you might have a patch adding "support for
instruction group foo" which includes everything needed
for those instructions including the translate.c changes
and the helper functions that get called by the generated
code.

Remember to make sure that everything still compiles at every
step in the patch series.

-- PMM