amdgpu/addrlib: fix pixel index calculation of thick micro tiling
authorXavi Zhang <xavi.zhang@amd.com>
Tue, 12 May 2015 08:26:59 +0000 (04:26 -0400)
committerMarek Olšák <marek.olsak@amd.com>
Thu, 30 Mar 2017 12:44:33 +0000 (14:44 +0200)
src/amd/addrlib/core/addrlib1.cpp

index 81455f0..0be7bc1 100644 (file)
@@ -2913,8 +2913,8 @@ ADDR_E_RETURNCODE AddrLib1::ComputeMicroTileEquation(
                 break;
             case 64:
             case 128:
-                pixelBit[0] = y0;
-                pixelBit[1] = x0;
+                pixelBit[0] = x0;
+                pixelBit[1] = y0;
                 pixelBit[2] = z0;
                 pixelBit[3] = x1;
                 pixelBit[4] = y1;
@@ -3124,8 +3124,8 @@ UINT_32 AddrLib1::ComputePixelIndexWithinMicroTile(
                 break;
             case 64:
             case 128:
-                pixelBit0 = y0;
-                pixelBit1 = x0;
+                pixelBit0 = x0;
+                pixelBit1 = y0;
                 pixelBit2 = z0;
                 pixelBit3 = x1;
                 pixelBit4 = y1;