It has been a while since I've used CygWin/MinGW/MSYS, but I thought that CygWin had it's own GCC, and the MinGW/MSYS compilers were not supposed to be combined with the CygWin stack, for license and tech reasons. I could be wrong...
Another new option may be to try Windows new Linux subsystem. Unclear if that will help with OP's issue with Source Insight.
On 03/08/2017 02:03 AM, Heyi Guo wrote:
I tried cygwin this week and it seems to work, at least successfully building the fd binary. I've no idea of how to combine the use of cygwin, mingw32 and Linaro mingw32 cross compiler, for the latest cygwin can only install mingw64 packages. What I use is a aarch64 gcc from a porting repo of cygwin. Next we will do some functional test upon the image built on cygwin.
Thanks for all your advices.
Heyi Guo (Gary)
On 27 February 2017 at 16:57, Leif Lindholm leif.lindholm@linaro.org wrote:
Hi Heyi,
On Fri, Feb 24, 2017 at 04:24:57PM +0800, Heyi Guo wrote:
I can see Linaro is Linux oriented, but most of my colleagues are much
more
familiar with Windows and Source Insight, rather than Linux and
vim/emacs.
We may be happier and more effective if we can build aarch64 UEFI on
Windows
directly.
So how can we do that? Is there any detailed guideline?
I have no experience of this myself, but as an observer I would like to point out that the solution I would instinctively have tried (cygwin/mingw + Linaro toolchains for Windows) comes with non-obvious drawbacks. The attempts to translate various filesystem properties sometimes causes breakages.
If you have access to Windows 10, https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux seems like it would be much less likely to break things.
But if people otherwise use Windows-type workflows, it sounds more useful if a solution like the one proposed by blibbet can be made to work.
Regards,
Leif
Linaro-uefi mailing list Linaro-uefi@lists.linaro.org https://lists.linaro.org/mailman/listinfo/linaro-uefi