Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Aug 31, 2020

• Use vector as TPC tracks MCLabels instead of MCTruthContainer

  shahor02

  committed
  6c7865c

  Copy full SHA for 6c7865c

• remove mass and ROFRame, add PID and layers pattern in ITStracks

  shahor02

  committed
  096ec0e

  Copy full SHA for 096ec0e

Commits on Aug 29, 2020

• TPCClusterFinder: Allow propagating mc labels when running on GPU.

  

  fweig

  authored and

  davidrohr

  committed
  9191664

  Copy full SHA for 9191664

• TPCClusterFinder: Improve mc label performance.

  

  fweig

  authored and

  davidrohr

  committed
  b18206b

  Copy full SHA for b18206b

Commits on Aug 27, 2020

• GPU: Some preparation work to support GPU run-time-compilation in the future

  davidrohr

  committed
  82c6f7a

  Copy full SHA for 82c6f7a

• GPU: Use only a single lane for clusterizing on CPU (saves memory, and they are anyway processed serially)

  davidrohr

  committed
  29ed57b

  Copy full SHA for 29ed57b

• TPC: Don't move digit labels into owning pointer

  davidrohr

  committed
  c3c4576

  Copy full SHA for c3c4576

Commits on Aug 26, 2020

• GPU: Fix AliRoot compilation

  davidrohr

  committed
  0158130

  Copy full SHA for 0158130

• GPU: Use C++17 namespace syntax in more places

  davidrohr

  committed
  8211b2c

  Copy full SHA for 8211b2c

• GPU: Use C++17 in CUDA compilation with CUDA 11/CMake 3.18

  davidrohr

  committed
  63e6af7

  Copy full SHA for 63e6af7

• GPU: Read missing setting that was lost when switching to configKeyValues

  

  davidrohr

  authored and

  ktf

  committed
  1724192

  Copy full SHA for 1724192

• GPU: Build standalone visualization only when DebugGUI present

  

  davidrohr

  authored and

  ktf

  committed
  07587c0

  Copy full SHA for 07587c0

• GPU: Apply GPU global memory reuse also for individual host allocations

  davidrohr

  committed
  d5fee9e

  Copy full SHA for d5fee9e

• GPU: No need to reinitialize pointers that were not initialized in the first place

  davidrohr

  committed
  8292063

  Copy full SHA for 8292063

• GPU: Fix typo leading to memory corruption

  davidrohr

  committed
  a3cf4c0

  Copy full SHA for a3cf4c0

• TPCClusterFinder: Reduce memory usage on MC propagation.

  

  fweig

  authored and

  davidrohr

  committed
  6b695d9

  Copy full SHA for 6b695d9

• GPU: Reduce memory for TPC clusterer MC label attachment

  davidrohr

  committed
  86d6b3d

  Copy full SHA for 86d6b3d

• Fix new codechecker errors from clang10

  davidrohr

  committed
  6542611

  Copy full SHA for 6542611

• GPU: Disable some settings by default when incompatible to the configuration

  davidrohr

  committed
  ddd815e

  Copy full SHA for ddd815e

Commits on Aug 25, 2020

• GPU: Select fastest gather kernel depending on architecture

  davidrohr

  committed
  c381bd4

  Copy full SHA for c381bd4

• GPU: Fix codechecker warning

  davidrohr

  committed
  542f311

  Copy full SHA for 542f311

• TPCClusterFinder: Fix mc propagation with ZS enabled.

  

  fweig

  authored and

  davidrohr

  committed
  cb4c925

  Copy full SHA for cb4c925

• TPCClusterFinder: Improve performance for mc labels.

  

  fweig

  authored and

  davidrohr

  committed
  dc09cd0

  Copy full SHA for dc09cd0

• GPU: Improve gather kernel performance

  Show description for 2f162f4
  

  fweig

  authored and

  davidrohr

  committed
  2f162f4

  Copy full SHA for 2f162f4

• GPU: Switch to CUDA 11

  davidrohr

  committed
  1dad6ed

  Copy full SHA for 1dad6ed

• GPU: Adjust CMake for hip-clang

  davidrohr

  committed
  24bb8a1

  Copy full SHA for 24bb8a1

• GPU: Adapt to hip-clang runtime

  davidrohr

  committed
  63611f8

  Copy full SHA for 63611f8

• GPU: Don't obtain timer objects when they are not required

  davidrohr

  committed
  45fcc74

  Copy full SHA for 45fcc74

Commits on Aug 24, 2020

• GPU: Query device timers only when debug mode enabled

  davidrohr

  committed
  7072735

  Copy full SHA for 7072735

• GPU: Define HIP CXX flags globally and add gpu-rdc to ITS code (no longer default with hip-clang)

  davidrohr

  committed
  0643be8

  Copy full SHA for 0643be8

• GPU: Flush denormals to zero also for HIP

  davidrohr

  committed
  91224a7

  Copy full SHA for 91224a7

• GPU: Remove HIP workaround to remove -g CXX flag

  davidrohr

  committed
  0dcec95

  Copy full SHA for 0dcec95

• Revert AMD workaround: hip-clang reports correct constant memory size now

  davidrohr

  committed
  23e186b

  Copy full SHA for 23e186b

• AMD Workaround: workaround for memory allocation problem in rocThrust

  davidrohr

  committed
  a3cde36

  Copy full SHA for a3cde36

• GPU: Add option to override number of blocks derrived from launch bounds

  davidrohr

  committed
  e3a97ab

  Copy full SHA for e3a97ab