| Message ID | 87vcq7ngxo.fsf@firetop.home |
|---|---|
| State | New |
| Headers | show |
Am 26.11.2011 14:37, schrieb Richard Sandiford: > There's some dodgy application of De Morgan's law in the emulation > of the MIPS BC1ANY[24]F instructions: they end up branching only > if all CCs are false, rather than if one CC is. > > Tested on mips64-linux-gnu, where it fixes the GCC MIPS3D tests. > > Signed-off-by: Richard Sandiford <rdsandiford@googlemail.com> > --- > target-mips/translate.c | 8 ++++---- > 1 files changed, 4 insertions(+), 4 deletions(-) Ping! Digging out another unreviewed mips patch for 1.1. Andreas > diff --git a/target-mips/translate.c b/target-mips/translate.c > index ba45eb0..2b977b3 100644 > --- a/target-mips/translate.c > +++ b/target-mips/translate.c > @@ -6017,7 +6017,7 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, > TCGv_i32 t1 = tcg_temp_new_i32(); > tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); > - tcg_gen_nor_i32(t0, t0, t1); > + tcg_gen_nand_i32(t0, t0, t1); > tcg_temp_free_i32(t1); > tcg_gen_andi_i32(t0, t0, 1); > tcg_gen_extu_i32_tl(bcond, t0); > @@ -6041,11 +6041,11 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, > TCGv_i32 t1 = tcg_temp_new_i32(); > tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); > - tcg_gen_or_i32(t0, t0, t1); > + tcg_gen_and_i32(t0, t0, t1); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2)); > - tcg_gen_or_i32(t0, t0, t1); > + tcg_gen_and_i32(t0, t0, t1); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3)); > - tcg_gen_nor_i32(t0, t0, t1); > + tcg_gen_nand_i32(t0, t0, t1); > tcg_temp_free_i32(t1); > tcg_gen_andi_i32(t0, t0, 1); > tcg_gen_extu_i32_tl(bcond, t0);
On 04/18/2012 04:00 PM, Andreas Färber wrote: > Am 26.11.2011 14:37, schrieb Richard Sandiford: >> There's some dodgy application of De Morgan's law in the emulation >> of the MIPS BC1ANY[24]F instructions: they end up branching only >> if all CCs are false, rather than if one CC is. >> >> Tested on mips64-linux-gnu, where it fixes the GCC MIPS3D tests. >> >> Signed-off-by: Richard Sandiford <rdsandiford@googlemail.com> >> --- >> target-mips/translate.c | 8 ++++---- >> 1 files changed, 4 insertions(+), 4 deletions(-) > > Ping! Digging out another unreviewed mips patch for 1.1. > > Andreas > >> diff --git a/target-mips/translate.c b/target-mips/translate.c >> index ba45eb0..2b977b3 100644 >> --- a/target-mips/translate.c >> +++ b/target-mips/translate.c >> @@ -6017,7 +6017,7 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, >> TCGv_i32 t1 = tcg_temp_new_i32(); >> tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); >> - tcg_gen_nor_i32(t0, t0, t1); >> + tcg_gen_nand_i32(t0, t0, t1); >> tcg_temp_free_i32(t1); >> tcg_gen_andi_i32(t0, t0, 1); >> tcg_gen_extu_i32_tl(bcond, t0); >> @@ -6041,11 +6041,11 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, >> TCGv_i32 t1 = tcg_temp_new_i32(); >> tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); >> - tcg_gen_or_i32(t0, t0, t1); >> + tcg_gen_and_i32(t0, t0, t1); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2)); >> - tcg_gen_or_i32(t0, t0, t1); >> + tcg_gen_and_i32(t0, t0, t1); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3)); >> - tcg_gen_nor_i32(t0, t0, t1); >> + tcg_gen_nand_i32(t0, t0, t1); >> tcg_temp_free_i32(t1); >> tcg_gen_andi_i32(t0, t0, 1); >> tcg_gen_extu_i32_tl(bcond, t0); > Reviewed-by: Richard Henderson <rth@twiddle.net> r~
Am 26.11.2011 14:37, schrieb Richard Sandiford: > There's some dodgy application of De Morgan's law in the emulation > of the MIPS BC1ANY[24]F instructions: they end up branching only > if all CCs are false, rather than if one CC is. > > Tested on mips64-linux-gnu, where it fixes the GCC MIPS3D tests. > > Signed-off-by: Richard Sandiford <rdsandiford@googlemail.com> > --- > target-mips/translate.c | 8 ++++---- > 1 files changed, 4 insertions(+), 4 deletions(-) Ping! Patch has a Reviewed-by. Anthony or Blue, can you apply for 1.1? http://patchwork.ozlabs.org/patch/127798/ /-F > diff --git a/target-mips/translate.c b/target-mips/translate.c > index ba45eb0..2b977b3 100644 > --- a/target-mips/translate.c > +++ b/target-mips/translate.c > @@ -6017,7 +6017,7 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, > TCGv_i32 t1 = tcg_temp_new_i32(); > tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); > - tcg_gen_nor_i32(t0, t0, t1); > + tcg_gen_nand_i32(t0, t0, t1); > tcg_temp_free_i32(t1); > tcg_gen_andi_i32(t0, t0, 1); > tcg_gen_extu_i32_tl(bcond, t0); > @@ -6041,11 +6041,11 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, > TCGv_i32 t1 = tcg_temp_new_i32(); > tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); > - tcg_gen_or_i32(t0, t0, t1); > + tcg_gen_and_i32(t0, t0, t1); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2)); > - tcg_gen_or_i32(t0, t0, t1); > + tcg_gen_and_i32(t0, t0, t1); > tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3)); > - tcg_gen_nor_i32(t0, t0, t1); > + tcg_gen_nand_i32(t0, t0, t1); > tcg_temp_free_i32(t1); > tcg_gen_andi_i32(t0, t0, 1); > tcg_gen_extu_i32_tl(bcond, t0);
Thanks, applied. On Sun, May 13, 2012 at 6:39 PM, Andreas Färber <afaerber@suse.de> wrote: > Am 26.11.2011 14:37, schrieb Richard Sandiford: >> There's some dodgy application of De Morgan's law in the emulation >> of the MIPS BC1ANY[24]F instructions: they end up branching only >> if all CCs are false, rather than if one CC is. >> >> Tested on mips64-linux-gnu, where it fixes the GCC MIPS3D tests. >> >> Signed-off-by: Richard Sandiford <rdsandiford@googlemail.com> >> --- >> target-mips/translate.c | 8 ++++---- >> 1 files changed, 4 insertions(+), 4 deletions(-) > > Ping! Patch has a Reviewed-by. Anthony or Blue, can you apply for 1.1? > > http://patchwork.ozlabs.org/patch/127798/ > > /-F > >> diff --git a/target-mips/translate.c b/target-mips/translate.c >> index ba45eb0..2b977b3 100644 >> --- a/target-mips/translate.c >> +++ b/target-mips/translate.c >> @@ -6017,7 +6017,7 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, >> TCGv_i32 t1 = tcg_temp_new_i32(); >> tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); >> - tcg_gen_nor_i32(t0, t0, t1); >> + tcg_gen_nand_i32(t0, t0, t1); >> tcg_temp_free_i32(t1); >> tcg_gen_andi_i32(t0, t0, 1); >> tcg_gen_extu_i32_tl(bcond, t0); >> @@ -6041,11 +6041,11 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, >> TCGv_i32 t1 = tcg_temp_new_i32(); >> tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); >> - tcg_gen_or_i32(t0, t0, t1); >> + tcg_gen_and_i32(t0, t0, t1); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2)); >> - tcg_gen_or_i32(t0, t0, t1); >> + tcg_gen_and_i32(t0, t0, t1); >> tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3)); >> - tcg_gen_nor_i32(t0, t0, t1); >> + tcg_gen_nand_i32(t0, t0, t1); >> tcg_temp_free_i32(t1); >> tcg_gen_andi_i32(t0, t0, 1); >> tcg_gen_extu_i32_tl(bcond, t0); > > -- > SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany > GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
diff --git a/target-mips/translate.c b/target-mips/translate.c index ba45eb0..2b977b3 100644 --- a/target-mips/translate.c +++ b/target-mips/translate.c @@ -6017,7 +6017,7 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, TCGv_i32 t1 = tcg_temp_new_i32(); tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); - tcg_gen_nor_i32(t0, t0, t1); + tcg_gen_nand_i32(t0, t0, t1); tcg_temp_free_i32(t1); tcg_gen_andi_i32(t0, t0, 1); tcg_gen_extu_i32_tl(bcond, t0); @@ -6041,11 +6041,11 @@ static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op, TCGv_i32 t1 = tcg_temp_new_i32(); tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc)); tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1)); - tcg_gen_or_i32(t0, t0, t1); + tcg_gen_and_i32(t0, t0, t1); tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2)); - tcg_gen_or_i32(t0, t0, t1); + tcg_gen_and_i32(t0, t0, t1); tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3)); - tcg_gen_nor_i32(t0, t0, t1); + tcg_gen_nand_i32(t0, t0, t1); tcg_temp_free_i32(t1); tcg_gen_andi_i32(t0, t0, 1); tcg_gen_extu_i32_tl(bcond, t0);
There's some dodgy application of De Morgan's law in the emulation of the MIPS BC1ANY[24]F instructions: they end up branching only if all CCs are false, rather than if one CC is. Tested on mips64-linux-gnu, where it fixes the GCC MIPS3D tests. Signed-off-by: Richard Sandiford <rdsandiford@googlemail.com> --- target-mips/translate.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)