OpenCL* for CPU
Ask questions and share information on Intel® SDK for OpenCL™ Applications and OpenCL™ implementations for Intel® CPU.
This forum covers OpenCL* for CPU only. OpenCL* for GPU questions can be asked in the GPU Compute Software forum. Intel® FPGA SDK for OpenCL™ questions can be ask in the FPGA Intel® High Level Design forum.
1720 Discussions

Installing on Debian Testing


I ran "./ install" as instructed, but I get this error:

Intel Graphics Driver Installer for OpenCL on CentOS

Running on Generic distribution

The installer will download (unless they already exist):
* OpenCL user-mode driver (87MiB) 
* Linux 4.7 kernel (66MiB) 

Setup requires ~10GiB disk space, and takes ~1h on a 4-cores machine.
(~8.9GiB for building the Linux kernel, ~1GiB for other files)

Do you want to continue? [y/n] y
Found existing workspace at /root/intel-opencl...
~/intel-opencl /home/dandan
Installing user mode driver...
Downloading user mode driver... already exists, skipping download...
unzip -o
  inflating: intel-opencl-cpu-r5.0-63503.x86_64.rpm  
  inflating: intel-opencl-cpu-r5.0-63503.x86_64.tar.xz  
  inflating: intel-opencl-cpu-r5.0-63503.x86_64.tar.xz.sig  
  inflating: intel-opencl-devel-r5.0-63503.x86_64.rpm  
  inflating: intel-opencl-devel-r5.0-63503.x86_64.tar.xz  
  inflating: intel-opencl-devel-r5.0-63503.x86_64.tar.xz.sig  
  inflating: intel-opencl-r5.0-63503.x86_64.rpm  
  inflating: intel-opencl-r5.0-63503.x86_64.tar.xz  
  inflating: intel-opencl-r5.0-63503.x86_64.tar.xz.sig  
  inflating: vpg_ocl_linux_rpmdeb.public  
Looking for previously installed user-mode driver...
rm -rf /etc/
rm -rf /etc/OpenCL/vendors/intel.icd
rm -rf /etc/profile.d/
rm -rf /opt/intel/opencl
tar -C intel-opencl -Jxf intel-opencl-cpu-r5.0-63503.x86_64.tar.xz
tar -C intel-opencl -Jxf intel-opencl-devel-r5.0-63503.x86_64.tar.xz
tar -C intel-opencl -Jxf intel-opencl-r5.0-63503.x86_64.tar.xz
cp -R intel-opencl/* /
Downloading Linux kernel 4.7...
linux-4.7.tar.xz already exists, skipping download...
Installing Linux kernel 4.7...
linux-4.7 kernel directory exists, skipping extract...
~/intel-opencl/linux-4.7 ~/intel-opencl /home/dandan
patch -p1 < /opt/intel/opencl/kernel-4.7.patch
can't find file to patch at input line 19
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
|From 18c0bd9224ea2764698bc42ceba47f1be4ed7967 Mon Sep 17 00:00:00 2001
|From: =?UTF-8?q?Micha=C5=82=20Winiarski?= <>
|Date: Wed, 21 Oct 2015 13:11:24 +0200
|Subject: [PATCH 01/17] drm/i915: Add L3_LRA_1 Register to cmdparser whitelist
|MIME-Version: 1.0
|Content-Type: text/plain; charset=UTF-8
|Content-Transfer-Encoding: 8bit
|Change-Id: I4fba318061b175d481ea2f2f02593d70930daf23
|Signed-off-by: Michał Winiarski <>
| drivers/gpu/drm/i915/i915_cmd_parser.c |    1 +
| 1 files changed, 1 insertions(+), 0 deletions(-)
|diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
|index a337f33..85f0f2f 100644
|--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
|+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
File to patch: 

What should I enter at "FIle to patch"? As I realized, it said "Intel Graphics Driver Installer for OpenCL on CentOS", but I'm running on Debian, which is similar enough to Ubuntu. How can I fix this?



0 Kudos
2 Replies


Even I am also facing the same issue as li, dandan . How can I fix this?


0 Kudos

Hello AdityaS and DandanL,

Thanks for sharing your setup experience.


Per guidance on the driver page, Debian/Ubuntu distros are not supported for the SRB5.0 release however they are expected to function... *Edit Certain Ubuntu versions are expected to function... see the SRB5.0 setup page for details... other Debian distros are not validated* the example script attempts to walk through the steps to patch and install 4.7 kernel as well as delpoy the SRB5.0 runtime. You could consider performing the steps manually. Please see the pdf here and consider walking through the install steps manually: 


Linked from here:

Best starting guesses for your error: ensure you have consistent permissions for where scripts read/write from/to... also I'll recommend starting by wiping all previous temp assets and tarballs that may have been left over situated by an earlier run.




Note: If you try a linux 4.8 kernel or greater, you may not need to perform any kernel patching. See Ubuntu 16.04.2 OS with vanilla 4.8 Kernel  in the above pdf link. This may be easier.


Lastly, consider using the "NEO" driver located here, which doesn't imply kernel patches as part of it's setup procedure. This may be easiest.:





0 Kudos