Notice: We have migrated to GitLab launching 2024-05-01 see here: https://gitlab.rtems.org/

#3630 closed defect (fixed)

Build of rtems-tools fails with i686-w64-mingw32

Reported by: Markus Bernd Moessner Owned by: Chris Johns
Priority: normal Milestone: 5.1
Component: arch/arm Version: 5
Severity: normal Keywords:
Cc: Blocked By:
Blocking:

Description

Hi,

I am following

https://docs.rtems.org/branches/master/user/hosts/windows.html

to build a Windows Host toolchain on Linux Mint 19. However, the build of rtems-tools fails with: "unknown host: i686-w64-mingw32".

I can track the issue down to the function "check_options" in the wscript. The function expects a host called "mingw32" or "x86_64-w64-mingw32". My naive solution would be to simply extend the list with "i686-w64-mingw32", but I've just started with RTEMS so I might have choosen a wrong path in an earlier step.

Attachments (2)

rsb-report-rtems-tools-6db01e577fed1dc88018106b81dd531f2ecc1fd0-1.txt (54.5 KB) - added by Markus Bernd Moessner on 12/02/18 at 16:05:51.
RSB Report
0001-Allow-build-with-i686-w64-mingw32.patch (734 bytes) - added by Markus Bernd Moessner on 12/02/18 at 16:06:58.
Patch to allow build of rtems-tools with i686-w64-mingw32

Download all attachments as: .zip

Change History (9)

Changed on 12/02/18 at 16:05:51 by Markus Bernd Moessner

RSB Report

Changed on 12/02/18 at 16:06:58 by Markus Bernd Moessner

Patch to allow build of rtems-tools with i686-w64-mingw32

comment:1 Changed on 12/03/18 at 09:13:10 by Markus Bernd Moessner

Component: adminarch/arm

comment:2 Changed on 12/03/18 at 19:15:43 by Markus Bernd Moessner

Build works fine with a 64bit toolchain. The reason why I was using i686 was the example given in

https://docs.rtems.org/branches/master/rsb/cross-canadian-cross.html#cross-building

comment:3 Changed on 12/04/18 at 03:23:36 by Chris Johns

Milestone: 5.1
Owner: set to Chris Johns
Status: newaccepted

Thanks. I will update the documentation and will close the ticket when I commit the change.

comment:4 Changed on 12/06/18 at 21:00:30 by Markus Bernd Moessner

Dear Chris,

sorry for getting back on this, but the question that I have is: is it intendend / expected that the build fails with the i686 toolchain? If not - then i'll look into it and submit a patch. If it is expected - well yeah please update the documentation.

Thanks & Regards
Markus

comment:5 in reply to:  4 Changed on 12/09/18 at 23:41:50 by Chris Johns

Replying to Markus Bernd Moessner:

sorry for getting back on this, but the question that I have is: is it intendend / expected that the build fails with the i686 toolchain?

No it is not. It is a simple matter of there being limited time and many possible combinations.

If not - then i'll look into it and submit a patch. If it is expected - well yeah please update the documentation.

Updates are most welcome. Thank you.

comment:6 Changed on 12/10/18 at 16:57:22 by Markus Bernd Moessner

Ok, thank you for taking the time to clarify this.

comment:7 Changed on 04/28/20 at 01:45:51 by kaidoho <kho237115@…>

Resolution: fixed
Status: acceptedclosed

In [changeset:"a1a05c7c630529e90fab632c6750bb2e825538b6/rtems-tools" a1a05c7/rtems-tools]:

Error: Processor CommitTicketReference failed
/data/trac/repo/rtems-tools.git does not appear to be a Git repository.
Note: See TracTickets for help on using tickets.