Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-09 Thread Jan Beulich
>>> On 08.01.15 at 20:29, wrote: > On 08/01/15 15:50, Jan Beulich wrote: >> +if ( !buf ) >> +buf = p_data; >> +else >> +switch ( bytes_per_rep ) >> +{ >> +#define CASE(bits, suffix) \ >> +case (bit

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Andrew Cooper
On 08/01/15 15:50, Jan Beulich wrote: > +if ( !buf ) > +buf = p_data; > +else > +switch ( bytes_per_rep ) > +{ > +#define CASE(bits, suffix) \ > +case (bits) / 8:

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Tim Deegan
At 17:05 + on 08 Jan (1420733142), Jan Beulich wrote: > >>> On 08.01.15 at 17:56, wrote: > > At 16:29 + on 08 Jan (1420730974), Jan Beulich wrote: > >> >>> On 08.01.15 at 17:16, wrote: > >> > At 15:50 + on 08 Jan (1420728649), Jan Beulich wrote: > >> >> While the REP MOVS acceleration

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Jan Beulich
>>> On 08.01.15 at 17:56, wrote: > At 16:29 + on 08 Jan (1420730974), Jan Beulich wrote: >> >>> On 08.01.15 at 17:16, wrote: >> > At 15:50 + on 08 Jan (1420728649), Jan Beulich wrote: >> >> While the REP MOVS acceleration appears to have helped qemu-traditional >> >> based guests, qemu-up

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Tim Deegan
At 16:29 + on 08 Jan (1420730974), Jan Beulich wrote: > >>> On 08.01.15 at 17:16, wrote: > > At 15:50 + on 08 Jan (1420728649), Jan Beulich wrote: > >> While the REP MOVS acceleration appears to have helped qemu-traditional > >> based guests, qemu-upstream (or really the respective video B

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Jan Beulich
>>> On 08.01.15 at 17:16, wrote: > At 15:50 + on 08 Jan (1420728649), Jan Beulich wrote: >> While the REP MOVS acceleration appears to have helped qemu-traditional >> based guests, qemu-upstream (or really the respective video BIOSes) >> doesn't appear to benefit from that. Instead the acceler

Re: [Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Tim Deegan
At 15:50 + on 08 Jan (1420728649), Jan Beulich wrote: > While the REP MOVS acceleration appears to have helped qemu-traditional > based guests, qemu-upstream (or really the respective video BIOSes) > doesn't appear to benefit from that. Instead the acceleration added > here provides a visible p

[Xen-devel] [PATCH 1/3] x86: also allow REP STOS emulation acceleration

2015-01-08 Thread Jan Beulich
While the REP MOVS acceleration appears to have helped qemu-traditional based guests, qemu-upstream (or really the respective video BIOSes) doesn't appear to benefit from that. Instead the acceleration added here provides a visible performance improvement during very early HVM guest boot. Signed-o