On 15 August 2012 16:37, Kenneth Graunke <[email protected]> wrote:
> 16-wide programs use roughly twice as many registers as 8-wide, and we > don't support spilling in 16-wide. So if an 8-wide program uses more > than half the available GRFs, the 16-wide one almost certainly will fail > to compile during register allocation. > > Not only that, but attempting to compiling such shaders is expensive: > programs that use a lot of registers tend to be quite complex, meaning > that we spend more time than usual generating and optimizing code. If > we fail at register allocation, we've failed at the last step, after > needlessly burning through a lot of CPU time. > > To make things worse, such shader compilation typically happens at the > first draw call using the shader, so it can cause the GPU to stall. > > With all that in mind, it makes sense to short-circuit the 16-wide > attempt if the 8-wide program uses too many registers. I've chosen 75 > to be conservative---if we /can/ compile a SIMD16 program, we want to. > > Reduces the number of GPU stalls due to fragment shader recompiles > in Left 4 Dead 2 by about 20%, and reduces the duration of many of the > remaining stalls by about half. > > Signed-off-by: Kenneth Graunke <[email protected]> > --- > src/mesa/drivers/dri/i965/brw_fs.cpp | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp > b/src/mesa/drivers/dri/i965/brw_fs.cpp > index e2dafdc..a113105 100644 > --- a/src/mesa/drivers/dri/i965/brw_fs.cpp > +++ b/src/mesa/drivers/dri/i965/brw_fs.cpp > @@ -2100,7 +2100,10 @@ brw_wm_fs_emit(struct brw_context *brw, struct > brw_wm_compile *c, > return false; > } > > - if (intel->gen >= 5 && c->prog_data.nr_pull_params == 0) { > + if (v.grf_used >= 75) { > + perf_debug("Too many registers to attempt compiling a 16-wide > shader; " > + "falling back to 8-wide at a 10-20%% performance > cost.\n"); > + } else if (intel->gen >= 5 && c->prog_data.nr_pull_params == 0) { > It looks like this code will give the perf warning even for situations where we couldn't do 16-wide anyhow (intel->gen == 4 || c->prog_data.nr_pull_params != 0). To avoid confusing people, we should probably only give the perf warning if register count is the *only* reason we can't do 16-wide. > c->dispatch_width = 16; > fs_visitor v2(c, prog, shader); > v2.import_uniforms(&v); > -- > 1.7.11.4 > > _______________________________________________ > mesa-dev mailing list > [email protected] > http://lists.freedesktop.org/mailman/listinfo/mesa-dev >
_______________________________________________ mesa-dev mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/mesa-dev
