diff options
| author | Brecht Van Lommel <brechtvanlommel@gmail.com> | 2017-11-01 23:02:28 +0300 |
|---|---|---|
| committer | Brecht Van Lommel <brechtvanlommel@gmail.com> | 2017-11-05 22:48:33 +0300 |
| commit | 8a72be7697f8fbfc8cb6cc9f3df049104e41d4a6 (patch) | |
| tree | c8997adcc23053eb6b4cfb7f499581644b23f61e /intern/cycles/kernel/kernel_emission.h | |
| parent | c571be4e05788b8d3447a0bfe59942ebb4464750 (diff) | |
Cycles: reduce closure memory usage for emission/shadow shader data.
With a Titan Xp, reduces path trace local memory from 1092MB to 840MB.
Benchmark performance was within 1% with both RX 480 and Titan Xp.
Original patch was implemented by Sergey.
Differential Revision: https://developer.blender.org/D2249
Diffstat (limited to 'intern/cycles/kernel/kernel_emission.h')
| -rw-r--r-- | intern/cycles/kernel/kernel_emission.h | 7 |
1 files changed, 2 insertions, 5 deletions
diff --git a/intern/cycles/kernel/kernel_emission.h b/intern/cycles/kernel/kernel_emission.h index 45b8c6311e1..94b0a37ce62 100644 --- a/intern/cycles/kernel/kernel_emission.h +++ b/intern/cycles/kernel/kernel_emission.h @@ -70,14 +70,11 @@ ccl_device_noinline float3 direct_emissive_eval(KernelGlobals *kg, /* no path flag, we're evaluating this for all closures. that's weak but * we'd have to do multiple evaluations otherwise */ path_state_modify_bounce(state, true); - shader_eval_surface(kg, emission_sd, state, 0); + shader_eval_surface(kg, emission_sd, state, 0, 0); path_state_modify_bounce(state, false); /* evaluate emissive closure */ - if(emission_sd->flag & SD_EMISSION) - eval = shader_emissive_eval(kg, emission_sd); - else - eval = make_float3(0.0f, 0.0f, 0.0f); + eval = shader_emissive_eval(kg, emission_sd); } eval *= ls->eval_fac; |