Properly bind shadow maps when stereo texture arrays are being used

Using LightweightPipeline.SetRenderTarget unilaterally considers intermediate targets as being texture arrays, which is not true for the shadow maps.

ScriptableRenderPipeline/LightweightPipeline/LWRP/LightweightPipeline.cs

             var cmd = CommandBufferPool.Get("Prepare Shadowmap");
             cmd.GetTemporaryRT(m_ShadowMapRTID, m_ShadowSettings.shadowAtlasWidth,
                 m_ShadowSettings.shadowAtlasHeight, kDepthStencilBufferBits, FilterMode.Bilinear, m_ShadowSettings.renderTextureFormat);
-            SetRenderTarget(cmd, m_ShadowMapRT, ClearFlag.Depth);
+            // LightweightPipeline.SetRenderTarget is meant to be used with camera targets, not shadowmaps
+            CoreUtils.SetRenderTarget(cmd, m_ShadowMapRT, ClearFlag.Depth, CoreUtils.ConvertSRGBToActiveColorSpace(m_CurrCamera.backgroundColor));
 
             if (shadowLight.lightType == LightType.Spot)
             {