[QUOTE=ewmayer;471979]Did your above unthreaded build attempt use the modified has_asimd() I posted? If it did and you got no errors in the compilation of util.c, that means you can go back to your original -DUSE_THREAD build (assuming you did that in a separate obj-file directory, and the resulting .o files are still around) and simply do an incremental recompile of util.c, and that should allow you link an executable. If you did your latest build in the same obj-file directory, I suggest you create a 2nd dir strictly for threaded-build obj-files, cd to that, then first make sure compilation of util.c works, then retry the all-sourcefiles compile.[/QUOTE]

Yes, I used the modified version in the unthreaded build. The threaded build succeeded and works (ran the test command and got the right output)

Will run a self test now.

Thanks for your help



Out of interest; is the source for this available on GitHub or some other open source platform?

[QUOTE=BrainStone;472082]Yes, I used the modified version in the unthreaded build. The threaded build succeeded and works (ran the test command and got the right output)

Will run a self test now.

Thanks for your help[/QUOTE]
Glad to hear it, and happy to do it - I think I fixed the issues that caused your unthreaded build to fail, will bundle up the modified files and post here for you to try in a few hours.

[QUOTE]Out of interest; is the source for this available on GitHub or some other open source platform?[/QUOTE]

No, the mersenneforum.org-linked is the only official ftp site for the code. I briefly flirted with Souceforge some years back, but as with George and his Prime95 client, this has always been pretty much a 1-man show, though I've had plenty of folks send me suggestions and their own local code mods to include in my upcoming releases. You are welcome to do the same.

[QUOTE=ewmayer;472089]No, the mersenneforum.org-linked is the only official ftp site for the code. I briefly flirted with Souceforge some years back, but as with George and his Prime95 client, this has always been pretty much a 1-man show, though I've had plenty of folks send me suggestions and their own local code mods to include in my upcoming releases. You are welcome to do the same.[/QUOTE]

Well GitHub is a great platform. And git (or some other scm) should be used to develop software anyways :P

[QUOTE=ewmayer;472089]I think I fixed the issues that caused your unthreaded build to fail, will bundle up the modified files and post here for you to try in a few hours.[/QUOTE]

AFAICT, 2 files need patches to fix the errors in your unthreaded build - I think it's a good thing for my builders and power-users to get their hands a little dirty, as it were, so the first of these is via simple edit to your local sourcefile. :) In radix144_ditN_cy_dif1.c, swap the following two lines, 180 and 181 (i.e. move the 'double' decl. outside the #if-block):
[code] double wt_re,wt_im, wi_re,wi_im; // Fermat-mod/LOACC weights stuff, used in both scalar and SIMD mode
#endif[/code]
And here is the patched radix52_ditN_cy_dif1.c - if incremental-rebuild of both files (gcc -c -O3 [path-to-src-dir]/radix52_ditN_cy_dif1.c, followed by ^52^144) succeeds, try linking, and then you'll simply have to wait for your threaded-build's self-tests to finish before running same using the unthreaded build: