X-Git-Url: https://git.cworth.org/git?a=blobdiff_plain;f=TODO;h=e7f4276772d3fe8f1d3700721d51b157825162e8;hb=07bf218026f6921c0b62714df9ec441b24884537;hp=f21f05230d68b1f0d13ffc405d1a67cd103467c7;hpb=6bfaad321892191c4e429fe302a819dff92cdca6;p=fips

diff --git a/TODO b/TODO
index f21f052..e7f4276 100644
--- a/TODO
+++ b/TODO
@@ -63,3 +63,17 @@ Explore using perf/LTTng probing instead of LD_PRELOAD wrapper
 
 	This has the advantage of allowing full-system,
 	multi-process data collection.
+
+Things to change while replaying to isolate impact of different pieces
+of the pipeline (list by Eero)
+==============================
+    The interesting "what is a bottleneck" experiments are
+    (from back of 3D pipeline to front):
+    - Disable blend (= disable related dst reads)
+    - Binding 1x1 / NULL texture instead of real ones [1]
+    - Using simple pass-through pixel shader instead of real ones [2]
+    - Use kill / discard pixel shader instead of real ones
+    - Use 0x0 / 1x1 scissor rect
+    - Front+backface cull
+    - Disable draws
+