From c7b1733105f803ea54e60432060f2c7d759221bc Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Luk=C3=A1=C5=A1=20Krup=C4=8D=C3=ADk?= <lukas.krupcik@vsb.cz>
Date: Mon, 21 Sep 2020 12:26:49 +0200
Subject: [PATCH 1/4] add intel-vtune-profiler
---
.../debuggers/intel-vtune-profiler.md | 70 +++++++++++++++++++
mkdocs.yml | 1 +
2 files changed, 71 insertions(+)
create mode 100644 docs.it4i/software/debuggers/intel-vtune-profiler.md
diff --git a/docs.it4i/software/debuggers/intel-vtune-profiler.md b/docs.it4i/software/debuggers/intel-vtune-profiler.md
new file mode 100644
index 000000000..9a2cf4dc9
--- /dev/null
+++ b/docs.it4i/software/debuggers/intel-vtune-profiler.md
@@ -0,0 +1,70 @@
+# Intel VTune Profiler
+
+## Introduction
+
+Intel ® VTune™ Profiler, part of Intel Parallel studio, is a GUI profiling tool designed for Intel processors. It offers a graphical performance analysis of single core and multithreaded applications. A highlight of the features:
+
+* Hotspot analysis
+* Locks and waits analysis
+* Low level specific counters, such as branch analysis and memory bandwidth
+* Power usage analysis - frequency and sleep states.
+
+## Usage
+
+To profile an application with VTune Profiler, special kernel modules need to be loaded. The modules are not loaded on the login nodes, thus direct profiling on login nodes is not possible. By default, the kernel modules ale not loaded on compute nodes neither. In order to have the modules loaded, you need to specify vtune=version PBS resource at job submit. The version is the same as for environment module. For example to use VTune/2020_update2-GCC:
+
+```console
+$ qsub -q qexp -A OPEN-0-0 -I -l select=1,vtune=2020_update2-GCC
+```
+
+After that, you can verify the modules `sep*`, `pax` and `vtsspp` are present in the kernel :
+
+```console
+login@cn191:~$ lsmod | grep -e sep -e pax -e vtsspp
+vtsspp 367418 0
+sep5 885333 0
+socperf3 595104 2 sep5,socwatch2_12
+pax 13820 0
+```
+
+To launch the GUI, first load the module:
+
+```console
+$ ml VTune/2020_update2-GCC
+```
+
+and launch the GUI :
+
+```console
+$ vtune-gui
+```
+
+!!! Warning
+ The command line `amplxe-gui` is deprecated. Use `vtune-gui` instead.
+
+The GUI will open in a new window. Click on "New Project..." to create a new project. After clicking OK, a new window with project properties will appear. At "Application:", select the bath to your binary you want to profile (the binary should be compiled with -g flag). Some additional options such as command line arguments can be selected. After clicking OK, your project is created.
+
+To run a new analysis, click "New analysis...". You will see a list of possible analyses. Some of them will not be possible on the current CPU, the GUI will show an error box if you select the wrong analysis. Clicking on Start will start profiling of the application.
+
+## Remote Analysis
+
+VTune Profiler also allows a form of remote analysis. In this mode, data for analysis is collected from the command line without GUI, and the results are then loaded to GUI on another machine. This allows profiling without interactive graphical jobs. To perform a remote analysis, launch a GUI somewhere, open the new analysis window and then click the button "Command line" in bottom right corner. It will show the command line needed to perform the selected analysis.
+
+The command line will look like this:
+
+```console
+/apps/all/VTune/2020_update2-GCC/vtune_profiler_2020.2.0.610396/bin64/vtune -collect hotspots -app-working-dir /home/$USER/tmp -- /home/$USER/tmp/sgemm
+```
+
+!!! Warning
+ The command line `amplxe-cl` is relative link to command `vtune`.
+
+Copy the line to clipboard and then you can paste it in your jobscript or in command line. After the collection is run, open the GUI once again, click the menu button in the upper right corner, and select "Open > Result...". The GUI will load the results from the run.
+
+## References
+
+1. [Intel® VTune™ Amplifier Support][a]
+1. [Amplifier Help Linux][b]
+
+[a]: https://software.intel.com/en-us/intel-vtune-amplifier-xe-support/documentation
+[b]: https://software.intel.com/en-us/amplifier_help_linux
diff --git a/mkdocs.yml b/mkdocs.yml
index b740e0b3d..a66a3c9f9 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -127,6 +127,7 @@ nav:
- CUBE: software/debuggers/cube.md
- Intel Performance Counter Monitor: software/debuggers/intel-performance-counter-monitor.md
- Intel VTune Amplifier XE: software/debuggers/intel-vtune-amplifier.md
+ - Intel VTune Profiler: software/debuggers/intel-vtune-profiler.md
- PAPI: software/debuggers/papi.md
- Scalasca: software/debuggers/scalasca.md
- Score-P: software/debuggers/score-p.md
--
GitLab
From 00683cf975ef5d90b8910ae562dbebad081c5b33 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Luk=C3=A1=C5=A1=20Krup=C4=8D=C3=ADk?= <lukas.krupcik@vsb.cz>
Date: Mon, 21 Sep 2020 12:34:21 +0200
Subject: [PATCH 2/4] typo
---
docs.it4i/anselm/storage.md | 2 +-
docs.it4i/barbora/storage.md | 2 +-
.../salomon/software/numerical-libraries/Clp.md | 2 +-
.../debuggers/intel-performance-counter-monitor.md | 14 +++++++-------
4 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/docs.it4i/anselm/storage.md b/docs.it4i/anselm/storage.md
index 752bffc99..488f69074 100644
--- a/docs.it4i/anselm/storage.md
+++ b/docs.it4i/anselm/storage.md
@@ -297,7 +297,7 @@ The local RAM disk filesystem is intended for temporary scratch data generated d
| Throughput | over 1.5 GB/s write, over 5 GB/s read, single thread, over 10 GB/s write, over 50 GB/s read, 16 threads |
| User quota | none |
-### Tmp
+### TMP
Each node is equipped with local /tmp directory of few GB capacity. The /tmp directory should be used to work with small temporary files. Old files in /tmp directory are automatically purged.
diff --git a/docs.it4i/barbora/storage.md b/docs.it4i/barbora/storage.md
index 9aa4adfe0..77ff3b0dd 100644
--- a/docs.it4i/barbora/storage.md
+++ b/docs.it4i/barbora/storage.md
@@ -245,7 +245,7 @@ Default ACL mechanism can be used to replace setuid/setgid permissions on direct
## Local Filesystems
-### Tmp
+### TMP
Each node is equipped with local /tmp directory of few GB capacity. The /tmp directory should be used to work with small temporary files. Old files in /tmp directory are automatically purged.
diff --git a/docs.it4i/salomon/software/numerical-libraries/Clp.md b/docs.it4i/salomon/software/numerical-libraries/Clp.md
index 5f2335e54..53dc5f4c2 100644
--- a/docs.it4i/salomon/software/numerical-libraries/Clp.md
+++ b/docs.it4i/salomon/software/numerical-libraries/Clp.md
@@ -1,4 +1,4 @@
-# Clp
+# CLP
## Introduction
diff --git a/docs.it4i/software/debuggers/intel-performance-counter-monitor.md b/docs.it4i/software/debuggers/intel-performance-counter-monitor.md
index 3ed10b73f..fac0f610a 100644
--- a/docs.it4i/software/debuggers/intel-performance-counter-monitor.md
+++ b/docs.it4i/software/debuggers/intel-performance-counter-monitor.md
@@ -16,7 +16,7 @@ $ ml intelpcm
PCM provides a set of tools to monitor system/or application.
-### Pcm-Memory
+### PCM-Memory
Measures memory bandwidth of your application or the whole system. Usage:
@@ -55,23 +55,23 @@ Sample output:
---------------------------------------||---------------------------------------
```
-### Pcm-Msr
+### PCM-Msr
Command pcm-msr.x can be used to read/write model specific registers of the CPU.
-### Pcm-Numa
+### PCM-Numa
NUMA monitoring utility does not work on Anselm.
-### Pcm-Pcie
+### PCM-Pcie
Can be used to monitor PCI Express bandwith. Usage: pcm-pcie.x <delay>
-### Pcm-Power
+### PCM-Power
Displays energy usage and thermal headroom for CPU and DRAM sockets. Usage: `pcm-power.x <delay> | <external program>`
-### Pcm
+### PCM
This command provides an overview of performance counters and memory usage. Usage: `pcm.x <delay> | <external program>`
@@ -184,7 +184,7 @@ Sample output :
Cleaning up
```
-### Pcm-Sensor
+### PCM-Sensor
Can be used as a sensor for ksysguard GUI, which is currently not installed on Anselm.
--
GitLab
From c2e798fbddcbbf59321fb5b8d2ae6b525307a493 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Luk=C3=A1=C5=A1=20Krup=C4=8D=C3=ADk?= <lukas.krupcik@vsb.cz>
Date: Mon, 21 Sep 2020 12:36:12 +0200
Subject: [PATCH 3/4] MD009 Trailing spaces
---
docs.it4i/software/debuggers/intel-vtune-profiler.md | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs.it4i/software/debuggers/intel-vtune-profiler.md b/docs.it4i/software/debuggers/intel-vtune-profiler.md
index 9a2cf4dc9..49a0ad516 100644
--- a/docs.it4i/software/debuggers/intel-vtune-profiler.md
+++ b/docs.it4i/software/debuggers/intel-vtune-profiler.md
@@ -1,4 +1,4 @@
-# Intel VTune Profiler
+# Intel VTune Profiler
## Introduction
--
GitLab
From d1a347bf05f62bdb5f589790dad39abe266e9704 Mon Sep 17 00:00:00 2001
From: Jan Siwiec <jan.siwiec@vsb.cz>
Date: Mon, 21 Sep 2020 12:53:22 +0200
Subject: [PATCH 4/4] proofreading intel-vtune-profiler.md
---
.../software/debuggers/intel-vtune-profiler.md | 18 +++++++++---------
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/docs.it4i/software/debuggers/intel-vtune-profiler.md b/docs.it4i/software/debuggers/intel-vtune-profiler.md
index 49a0ad516..ce632b6cc 100644
--- a/docs.it4i/software/debuggers/intel-vtune-profiler.md
+++ b/docs.it4i/software/debuggers/intel-vtune-profiler.md
@@ -2,7 +2,7 @@
## Introduction
-Intel ® VTune™ Profiler, part of Intel Parallel studio, is a GUI profiling tool designed for Intel processors. It offers a graphical performance analysis of single core and multithreaded applications. A highlight of the features:
+Intel® VTune™ Profiler, part of Intel Parallel studio, is a GUI profiling tool designed for Intel processors. It offers a graphical performance analysis of single-core and multithreaded applications. A highlight of the features:
* Hotspot analysis
* Locks and waits analysis
@@ -11,13 +11,13 @@ Intel ® VTune™ Profiler, part of Intel Parallel studio, is a GUI profiling to
## Usage
-To profile an application with VTune Profiler, special kernel modules need to be loaded. The modules are not loaded on the login nodes, thus direct profiling on login nodes is not possible. By default, the kernel modules ale not loaded on compute nodes neither. In order to have the modules loaded, you need to specify vtune=version PBS resource at job submit. The version is the same as for environment module. For example to use VTune/2020_update2-GCC:
+To profile an application with VTune Profiler, special kernel modules need to be loaded. The modules are not loaded on the login nodes, thus direct profiling on login nodes is not possible. By default, the kernel modules are not loaded on compute nodes either. In order to have the modules loaded, you need to specify the `vtune=version` PBS resource at job submit. The version is the same as for the environment module. For example, to use VTune/2020_update2-GCC:
```console
$ qsub -q qexp -A OPEN-0-0 -I -l select=1,vtune=2020_update2-GCC
```
-After that, you can verify the modules `sep*`, `pax` and `vtsspp` are present in the kernel :
+After that, you can verify that the modules `sep*`, `pax`, and `vtsspp` are present in the kernel:
```console
login@cn191:~$ lsmod | grep -e sep -e pax -e vtsspp
@@ -33,7 +33,7 @@ To launch the GUI, first load the module:
$ ml VTune/2020_update2-GCC
```
-and launch the GUI :
+and launch the GUI:
```console
$ vtune-gui
@@ -42,13 +42,13 @@ $ vtune-gui
!!! Warning
The command line `amplxe-gui` is deprecated. Use `vtune-gui` instead.
-The GUI will open in a new window. Click on "New Project..." to create a new project. After clicking OK, a new window with project properties will appear. At "Application:", select the bath to your binary you want to profile (the binary should be compiled with -g flag). Some additional options such as command line arguments can be selected. After clicking OK, your project is created.
+The GUI will open in a new window. Click on "New Project..." to create a new project. After clicking OK, a new window with project properties will appear. At "Application:", select the path to your binary you want to profile (the binary should be compiled with the `-g` flag). You can also select some additional options such as command line arguments. Click OK to create the project.
-To run a new analysis, click "New analysis...". You will see a list of possible analyses. Some of them will not be possible on the current CPU, the GUI will show an error box if you select the wrong analysis. Clicking on Start will start profiling of the application.
+To run a new analysis, click "New analysis...". You will see a list of possible analyses. Some of them will not be possible on the current CPU, the GUI will show an error box if you select a wrong analysis. Clicking on Start will start profiling of the application.
## Remote Analysis
-VTune Profiler also allows a form of remote analysis. In this mode, data for analysis is collected from the command line without GUI, and the results are then loaded to GUI on another machine. This allows profiling without interactive graphical jobs. To perform a remote analysis, launch a GUI somewhere, open the new analysis window and then click the button "Command line" in bottom right corner. It will show the command line needed to perform the selected analysis.
+VTune Profiler also allows a form of remote analysis. In this mode, data for analysis is collected from the command line without GUI, and the results are then loaded to GUI on another machine. This allows profiling without interactive graphical jobs. To perform a remote analysis, launch a GUI somewhere, open the new analysis window and then click the "Command line" button in the bottom right corner. It will show the command line needed to perform the selected analysis.
The command line will look like this:
@@ -57,9 +57,9 @@ The command line will look like this:
```
!!! Warning
- The command line `amplxe-cl` is relative link to command `vtune`.
+ The command line `amplxe-cl` is a relative link to command `vtune`.
-Copy the line to clipboard and then you can paste it in your jobscript or in command line. After the collection is run, open the GUI once again, click the menu button in the upper right corner, and select "Open > Result...". The GUI will load the results from the run.
+Copy the line to clipboard and then you can paste it in your jobscript or in the command line. After the collection is run, open the GUI again, click the menu button in the upper right corner, and select "Open > Result...". The GUI will load the results from the run.
## References
--
GitLab