On 03/05/2021 01:54, David Gibson wrote: > On Fri, Apr 30, 2021 at 04:35:28PM -0300, Bruno Larsen (billionai) wrote: >> Moved all SPR read/write callback, and some related functions, to a >> new file specific for it. These callbacks are TCG only, so separating >> them is required to support the build flag disable-tcg. >> >> Making the spr_noaccess function not static, and moving the define to >> internal.h is required, otherwise the project doesn't build >> >> Signed-off-by: Bruno Larsen (billionai) >> --- >> target/ppc/internal.h | 3 + >> target/ppc/spr_tcg.c.inc | 1052 +++++++++++++++++++++++++++++++ >> target/ppc/translate.c | 47 +- >> target/ppc/translate_init.c.inc | 981 ---------------------------- >> 4 files changed, 1056 insertions(+), 1027 deletions(-) >> create mode 100644 target/ppc/spr_tcg.c.inc > Hmm. So, I'm still pretty ambivalent about the creation of > spr_tcg.c.inc. The separation between it and translate.c just seems > rather arbitrary to me. > > Yes, translate.c is overly huge, but I'm not sure going from > overly-huge to overly-huger really makes anything worse, and would > avoid some of the dancing around to handle the new file. > > So for now, I'm inclined to suggest just moving the spr callbacks from > translate_init.c.inc to translate.c. And some day, I hope someone has > time to split up translate.c based on CPU family, which I think is a > more useful way to make it less huge. Ok. I agree that the improved is pretty minor, I was just thinking it was probably better than nothing... I'll work on that tomorrow and will have a new version. When I do, should I put the many e-mail IDs that hold required patches as "Based-on" tags, or should I just say ppc-for-6.1? I'm still a bit lost when dealing with patchew and these things... -- Bruno Piazera Larsen Instituto de Pesquisas ELDORADO Departamento Computação Embarcada Analista de Software Trainee Aviso Legal - Disclaimer