On 6/18/21 10:15 PM, Richard Henderson wrote: > On 6/17/21 10:49 AM, Philippe Mathieu-Daudé wrote: >> To be able to extract the microMIPS ISA and Code Compaction ASE >> translation routines to different source files, declare few TCG >> helpers which are also used by translate.c in "translate.h". >> >> Signed-off-by: Philippe Mathieu-Daudé<f4...@amsat.org> >> --- >> target/mips/tcg/translate.h | 5 +++++ >> target/mips/tcg/translate.c | 9 ++++----- >> 2 files changed, 9 insertions(+), 5 deletions(-) > > What patch set does this belong with?
micromips_translate.c.inc:1778: gen_ldxs(ctx, rs, rt, rd); micromips_translate.c.inc:1806: gen_align(ctx, 32, rd, rs, rt, extract32(ctx->opcode, 9, 2)); micromips_translate.c.inc:2859: gen_addiupc(ctx, reg, offset, 0, 0); mips16e_translate.c.inc:444: gen_addiupc(ctx, ry, offset, 1, extended); mips16e_translate.c.inc:481: gen_addiupc(ctx, rx, imm, 0, 1); mips16e_translate.c.inc:682: gen_addiupc(ctx, rx, ((uint8_t) ctx->opcode) << 2, 0, 0); > It doesn't seem to match the next > two patches, which don't create new compilation units. Indeed. In a yet-to-be-posted later series they get renamed to .c, becoming new units. It seemed simpler for me to do it that way, but I'll see how to rearrange the patches. Ah, now I remember, rearranging the patches modify the next 2 big patches (total 4k+ lines moved) you already reviewed... So to keep your two R-b it is simpler to add this one first. What I can do is hold these patches and post them later with the yet-to-be-posted series, *but* that breaks the "extract nanomips from translate.c" which you already reviewed :/ It is hard to split this huge 25k+ spaghetti translate.c :( > As far as it goes, the patch is fine... > Reviewed-by: Richard Henderson <richard.hender...@linaro.org> > > > r~ >
