Re: [Qemu-devel] [PATCH 1/2] target/mips: Improve performance for MSA binary operations

[Mateja Marjanovic <Mateja.Marjanovic@rt-rk.com>]

On 2.6.19. 09:06, Aleksandar Markovic wrote:
>
>
> On Jun 1, 2019 4:16 PM, "Aleksandar Markovic" <amarkovic@wavecomp.com 
> <mailto:amarkovic@wavecomp.com>> wrote:
> >
> > > From: Mateja Marjanovic <mateja.marjanovic@rt-rk.com 
> <mailto:mateja.marjanovic@rt-rk.com>>
> > > Sent: Monday, March 4, 2019 5:51 PM
> > > To: qemu-devel@nongnu.org <mailto:qemu-devel@nongnu.org>
> > > Cc: aurelien@aurel32.net <mailto:aurelien@aurel32.net>; Aleksandar 
> Markovic; Aleksandar Rikalo
> > > Subject: [PATCH 1/2] target/mips: Improve performance for MSA 
> binary operations
> > >
> > > From: Mateja Marjanovic <Mateja.Marjanovic@rt-rk.com 
> <mailto:Mateja.Marjanovic@rt-rk.com>>
> > >
> > > Eliminate loops for better performance.
> > >
> > > Signed-off-by: Mateja Marjanovic <mateja.marjanovic@rt-rk.com 
> <mailto:mateja.marjanovic@rt-rk.com>>
> > > ---
> > >  target/mips/msa_helper.c | 43 
> ++++++++++++++++++++++++++++++-------------
> > >  1 file changed, 30 insertions(+), 13 deletions(-)
> > >
> >
> > The commit message should be a little bit more informative - for 
> example,
> > it could list the affected instructions. Please consider other groups of
> > MSA instructions that are implemented via helpers that use similar "for"
> > loops. Otherwise:
> >
> > Reviewed-by: Aleksandar Markovic <amarkovic@wavecomp.com 
> <mailto:amarkovic@wavecomp.com>>
> >
>
> Mateja, you don't need to do anything regarding this patch, I am going 
> to fix the issues while appying.
>
Alright, thanks. :)

Regards,
Mateja
>
> Thanks, Aleksandar
>
> > > diff --git a/target/mips/msa_helper.c b/target/mips/msa_helper.c
> > > index 4c7ec05..1152fda 100644
> > > --- a/target/mips/msa_helper.c
> > > +++ b/target/mips/msa_helper.c
> > > @@ -804,28 +804,45 @@ void helper_msa_ ## func ## _df(CPUMIPSState 
> *env, uint32_t > df,         \
> > >      wr_t *pwd = &(env->active_fpu.fpr[wd].wr);                    
>       \
> > >      wr_t *pws = &(env->active_fpu.fpr[ws].wr);                    
>       \
> > >      wr_t *pwt = &(env->active_fpu.fpr[wt].wr);                    
>       \
> > > -    uint32_t i;                    \
> > >                   \
> > >      switch (df) {                    \
> > >      case DF_BYTE:                    \
> > > -        for (i = 0; i < DF_ELEMENTS(DF_BYTE); i++) {              
>       \
> > > -            pwd->b[i] = msa_ ## func ## _df(df, pws->b[i], 
> pwt->b[i]);  \
> > > -        }                    \
> > > +        pwd->b[0]  = msa_ ## func ## _df(df, pws->b[0], 
> pwt->b[0]);     \
> > > +        pwd->b[1]  = msa_ ## func ## _df(df, pws->b[1], 
> pwt->b[1]);     \
> > > +        pwd->b[2]  = msa_ ## func ## _df(df, pws->b[2], 
> pwt->b[2]);     \
> > > +        pwd->b[3]  = msa_ ## func ## _df(df, pws->b[3], 
> pwt->b[3]);     \
> > > +        pwd->b[4]  = msa_ ## func ## _df(df, pws->b[4], 
> pwt->b[4]);     \
> > > +        pwd->b[5]  = msa_ ## func ## _df(df, pws->b[5], 
> pwt->b[5]);     \
> > > +        pwd->b[6]  = msa_ ## func ## _df(df, pws->b[6], 
> pwt->b[6]);     \
> > > +        pwd->b[7]  = msa_ ## func ## _df(df, pws->b[7], 
> pwt->b[7]);     \
> > > +        pwd->b[8]  = msa_ ## func ## _df(df, pws->b[8], 
> pwt->b[8]);     \
> > > +        pwd->b[9]  = msa_ ## func ## _df(df, pws->b[9], 
> pwt->b[9]);     \
> > > +        pwd->b[10] = msa_ ## func ## _df(df, pws->b[10], 
> pwt->b[10]);   \
> > > +        pwd->b[11] = msa_ ## func ## _df(df, pws->b[11], 
> pwt->b[11]);   \
> > > +        pwd->b[12] = msa_ ## func ## _df(df, pws->b[12], 
> pwt->b[12]);   \
> > > +        pwd->b[13] = msa_ ## func ## _df(df, pws->b[13], 
> pwt->b[13]);   \
> > > +        pwd->b[14] = msa_ ## func ## _df(df, pws->b[14], 
> pwt->b[14]);   \
> > > +        pwd->b[15] = msa_ ## func ## _df(df, pws->b[15], 
> pwt->b[15]);   \
> > >          break;                   \
> > >      case DF_HALF:                    \
> > > -        for (i = 0; i < DF_ELEMENTS(DF_HALF); i++) {              
>       \
> > > -            pwd->h[i] = msa_ ## func ## _df(df, pws->h[i], 
> pwt->h[i]);  \
> > > -        }                    \
> > > +        pwd->h[0] = msa_ ## func ## _df(df, pws->h[0], 
> pwt->h[0]);      \
> > > +        pwd->h[1] = msa_ ## func ## _df(df, pws->h[1], 
> pwt->h[1]);      \
> > > +        pwd->h[2] = msa_ ## func ## _df(df, pws->h[2], 
> pwt->h[2]);      \
> > > +        pwd->h[3] = msa_ ## func ## _df(df, pws->h[3], 
> pwt->h[3]);      \
> > > +        pwd->h[4] = msa_ ## func ## _df(df, pws->h[4], 
> pwt->h[4]);      \
> > > +        pwd->h[5] = msa_ ## func ## _df(df, pws->h[5], 
> pwt->h[5]);      \
> > > +        pwd->h[6] = msa_ ## func ## _df(df, pws->h[6], 
> pwt->h[6]);      \
> > > +        pwd->h[7] = msa_ ## func ## _df(df, pws->h[7], 
> pwt->h[7]);      \
> > >          break;                   \
> > >      case DF_WORD:                    \
> > > -        for (i = 0; i < DF_ELEMENTS(DF_WORD); i++) {              
>       \
> > > -            pwd->w[i] = msa_ ## func ## _df(df, pws->w[i], 
> pwt->w[i]);  \
> > > -        }                    \
> > > +        pwd->w[0] = msa_ ## func ## _df(df, pws->w[0], 
> pwt->w[0]);      \
> > > +        pwd->w[1] = msa_ ## func ## _df(df, pws->w[1], 
> pwt->w[1]);      \
> > > +        pwd->w[2] = msa_ ## func ## _df(df, pws->w[2], 
> pwt->w[2]);      \
> > > +        pwd->w[3] = msa_ ## func ## _df(df, pws->w[3], 
> pwt->w[3]);      \
> > >          break;                   \
> > >      case DF_DOUBLE:                    \
> > > -        for (i = 0; i < DF_ELEMENTS(DF_DOUBLE); i++) {            
>       \
> > > -            pwd->d[i] = msa_ ## func ## _df(df, pws->d[i], 
> pwt->d[i]);  \
> > > -        }                    \
> > > +        pwd->d[0] = msa_ ## func ## _df(df, pws->d[0], 
> pwt->d[0]);      \
> > > +        pwd->d[1] = msa_ ## func ## _df(df, pws->d[1], 
> pwt->d[1]);      \
> > >          break;                   \
> > >      default:                   \
> > >          assert(0);                   \
> > > --
> > > 2.7.4
> > >
> > >
>