search for: r302679

...through an X register: > > FastISel: > fadd d0, d1, d0 > str d0, [sp,#528] > GlobalISel: > fadd d0, d1, d0 > fmov x9, d0 > stur x9, [x29,#-48] > > > Good finding, I forgot to do stores in my previous fix. I’ll do them > shortly. > > > Should be fixed by r302679 > > > Thanks Quentin, > > That reduces the slow-down when enabling globalisel at -O0 from 13% (on > r302453) to 9.5% (on r302679) in my experiments. > The code size increase also reduces from just over 3% to 2.8%. > > Kristof > > > ______________________________...

...through an X register: > > FastISel: > fadd d0, d1, d0 > str d0, [sp,#528] > GlobalISel: > fadd d0, d1, d0 > fmov x9, d0 > stur x9, [x29,#-48] > > > Good finding, I forgot to do stores in my previous fix. I’ll do them > shortly. > > > Should be fixed by r302679 > > > Thanks Quentin, > > That reduces the slow-down when enabling globalisel at -O0 from 13% (on > r302453) to 9.5% (on r302679) in my experiments. > The code size increase also reduces from just over 3% to 2.8%. > > Kristof > > > ______________________________...

...t; GlobalISel: > >> fadd d0, d1, d0 > >> fmov x9, d0 > >> stur x9, [x29,#-48] > >> > >> > >> Good finding, I forgot to do stores in my previous fix. I’ll do them > >> shortly. > >> > >> > >> Should be fixed by r302679 > >> > >> > >> Thanks Quentin, > >> > >> That reduces the slow-down when enabling globalisel at -O0 from 13% (on > >> r302453) to 9.5% (on r302679) in my experiments. > >> The code size increase also reduces from just over 3% to 2.8%....

...x9, d0 >>> >> stur x9, [x29,#-48] >>> >> >>> >> >>> >> Good finding, I forgot to do stores in my previous fix. I’ll do them >>> >> shortly. >>> >> >>> >> >>> >> Should be fixed by r302679 >>> >> >>> >> >>> >> Thanks Quentin, >>> >> >>> >> That reduces the slow-down when enabling globalisel at -O0 from 13% (on >>> >> r302453) to 9.5% (on r302679) in my experiments. >>> >> The cod...

...through an X register: > > FastISel: > fadd d0, d1, d0 > str d0, [sp,#528] > GlobalISel: > fadd d0, d1, d0 > fmov x9, d0 > stur x9, [x29,#-48] > > > Good finding, I forgot to do stores in my previous fix. I’ll do them > shortly. > > > Should be fixed by r302679 > > > Thanks Quentin, > > That reduces the slow-down when enabling globalisel at -O0 from 13% (on > r302453) to 9.5% (on r302679) in my experiments. > The code size increase also reduces from just over 3% to 2.8%. > > Kristof > > > ______________________________...

...>>> GlobalISel: >>> fadd d0, d1, d0 >>> fmov x9, d0 >>> stur x9, [x29,#-48] >>> >>> >>> Good finding, I forgot to do stores in my previous fix. I’ll do them >>> shortly. >>> >>> >>> Should be fixed by r302679 >>> >>> >>> Thanks Quentin, >>> >>> That reduces the slow-down when enabling globalisel at -O0 from 13% (on >>> r302453) to 9.5% (on r302679) in my experiments. >>> The code size increase also reduces from just over 3% to 2.8%. >&gt...

...through an X register: >> FastISel: >> fadd d0, d1, d0 >> str d0, [sp,#528] >> GlobalISel: >> fadd d0, d1, d0 >> fmov x9, d0 >> stur x9, [x29,#-48] > > Good finding, I forgot to do stores in my previous fix. I’ll do them shortly. Should be fixed by r302679 > >> >> Other overall impression when comparing code generation: The code size increase is probably mainly due to creating constants always in the entry BB of the function. For functions with many constants, this leads to lots and lots of constant creation and then immediately spi...

...through an X register: > > FastISel: > fadd d0, d1, d0 > str d0, [sp,#528] > GlobalISel: > fadd d0, d1, d0 > fmov x9, d0 > stur x9, [x29,#-48] > > > Good finding, I forgot to do stores in my previous fix. I’ll do them > shortly. > > > Should be fixed by r302679 > > > Thanks Quentin, > > That reduces the slow-down when enabling globalisel at -O0 from 13% (on > r302453) to 9.5% (on r302679) in my experiments. > The code size increase also reduces from just over 3% to 2.8%. > > Kristof > > > ______________________________...

Great Quentin :). I've rerun the benchmarks comparing "-O0 -g" with "-O0 -g -mllvm -global-isel -mllvm -global-isel-abort=2" on r302453, on AArch64 Cortex-A57. I indeed see almost no moves between GPR and FPR registers anymore (see details below for where I still see some). On geomean, I see 13% slow down (down from 17% on my previous run). On geomean, code size increase