Once the framework is in place adding addtional compile flags like --enable-multilib is relatively easy.. see

Alyssa submitted this set of patches for glmark2 to fix/clarify glmark2's handling of fp16 overflows in a shader (which succeed if the driver doesn't handle fp16 and uses fp32 instead). See:

Hi Peter, Am 17.12.20 um 19:59 schrieb Peter: Thanks for the detail explanation of your solution. The problem I see is that your solution doesn't scale. What happens if your system have a second not

The proof of concept layer is http://layers.openembedded.org/layerindex/branch/master/layer/meta-exotic/ - https://github.com/peteasa/meta-exotic/wiki. I created a set of cross compilers for the

just got an interesting request from a colleague ... for reasons i will explain shortly, it's necessary to build both the normal and debug versions of a library, but for the debug version to be

Interesting. As you know (but for reference for other people) the baud rate of the rock pi 4's serial console is 1,500,000bps and not the usual 115,200bps that everything else uses. I have *always*

Looking at the config.log, I only see the standard Yocto Project (nativesdk) paths, as well as: --disable-asm --without-capabilities I'm building on 18.04.4 LTS, I think it's a fairly standard

<mark.hatle@...> wrote: does it have --enable-optimization switch on ? if so then its probing the build platform for features and detects XMM registers and enables them but

Note, this is still happening. Using meta-mingw w/ gatesgarth to build, and including libgcrypt I get the same failure. The failure resolves with "--disable-amd64-as-feature-detection". Looking at

Hi Khem, Am 09.12.20 um 19:23 schrieb Khem Raj: The cross-canadian changes the HOST variables and not the TARGET_ variables. I need to change the TARGET variables but keep the HOST variables. It

Is there a project policy reason that all these patches are posted in the main list instead of a dev list? Isn't there a list specifically for this kind of thing, to separate questions and info

Hi Peter, Am 09.12.20 um 14:32 schrieb Peter Saunderson: Thanks for the links but your solution looks very complex. I will try to improve the examples from Richard. Regards Stefan

This series requires to work with two OE patches which I've submitted to OE mailing list //Hongxu

Target arch only supports 32 bit arm and 32 bit x86, 64 bit arm and 64 bit x86. BSP (MACHINE) incluced in above archs should be supported. Signed-off-by: Hongxu Jia <hongxu.jia@...> ---

Split sub packages (libtensorflow-c libtensorflow-framework label-image label-image-lite python3-tensorflow), then user chould select the required one rather than install them as a

Refers begginers of upstream [1] [2] [1] https://www.tensorflow.org/install/pip [2] https://www.tensorflow.org/tutorials/quickstart/beginner Signed-off-by: Hongxu Jia <hongxu.jia@...> ---

Set BAZEL_JOBS and BAZEL_MEM with empty string to make bazel build without limitation Signed-off-by: Hongxu Jia <hongxu.jia@...> --- classes/bazel.bbclass | 9

- Move common config settings to bazel-base.bbclass - Improve bazel_get_flags to make each flag per line with comments - Use BAZEL_MEM rather than hardcoded for option --local_ram_resources -

Signed-off-by: Hongxu Jia <hongxu.jia@...> --- ...l-native_3.1.0.bb => bazel-native_3.7.1.bb} | 4 ++-- ...e-sysroot-dir-to-the-default-Bazel-to.patch | 12 ++++++------

CROSSTOOL files are no longer supported after Bazel v0.26+ Signed-off-by: Hongxu Jia <hongxu.jia@...> --- .../tensorflow/files/CROSSTOOL.tpl | 230 ------------------