Brian Paul wrote: > > It'd be interesting to know how big the batch buffers are with and > without this change but overall this sounds good to me.
The attached graph shows the frequency of batch sizes for a 5 second run of glxgears on one SPU with original and modified batch packing[1]. Numbers were collected by printing the batch size before the start of each command processing loop in spu_command.c:cmd_batch(). I'm not sure that there is much that may be gleaned from this graph. Perhaps : 1. Batch sizes appear to be, on average, larger using the modified version. 2. This particular graph may not be the best way to represent the gathered data. 3. I like graphs. jonathan. [1] Something like: CELL_NUM_SPUS=1 LD_LIBRARY_PATH=$MESAGIT/lib glxgears > modified_list & sleep 5 && killall glxgears
<<inline: batch_size_freq.png>>
------------------------------------------------------------------------------ Check out the new SourceForge.net Marketplace. It is the best place to buy or sell services for just about anything Open Source. http://p.sf.net/sfu/Xq1LFB
_______________________________________________ Mesa3d-dev mailing list Mesa3d-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mesa3d-dev
