https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #11 from Jakub Jelinek ---
(In reply to Tom de Vries from comment #7)
> I think fexceptions is not implemented for nvptx. setjmp/longjmp is not
> supported on nvptx, so using that variant is not possible. The other
> possibility is us
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #10 from Jakub Jelinek ---
Indeed, lowering ignores map clauses for declare target to global vars, but it
can do that only if always is not present. Will have a look.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #9 from cesar at gcc dot gnu.org ---
(In reply to Tom de Vries from comment #8)
> Hmm, for me that minimizes to:
> ...
> extern "C" void abort ();
>
> #pragma omp declare target
> int results[2000];
> #pragma omp end declare target
>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #8 from Tom de Vries ---
(In reply to Tom de Vries from comment #7)
> (In reply to Jakub Jelinek from comment #6)
> > LLSDA labels are for EH support, this testcase doesn't really use exceptions
> > in any way, but it isn't compiled w
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #7 from Tom de Vries ---
(In reply to Jakub Jelinek from comment #6)
> LLSDA labels are for EH support, this testcase doesn't really use exceptions
> in any way, but it isn't compiled with -fno-exceptions; but that is the case
> of mo
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #6 from Jakub Jelinek ---
LLSDA labels are for EH support, this testcase doesn't really use exceptions in
any way, but it isn't compiled with -fno-exceptions; but that is the case of
most of the C++ tests in libgomp (I see only 6 Open
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #5 from Jakub Jelinek ---
So a nvptx backend bug?
Does ptxas require only decimal constants and not hexadecimal, or require
64-bit load from something that needs to be initialized with 64-bit initializer
rather than series of bytes, s
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #4 from Tom de Vries ---
(In reply to Tom de Vries from comment #3)
> (In reply to Jakub Jelinek from comment #2)
> > Created attachment 44437 [details]
> > gcc9-pr86660.patch
> >
> > Untested patch. Does this fix it?
>
> I'll try
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #3 from Tom de Vries ---
(In reply to Jakub Jelinek from comment #2)
> Created attachment 44437 [details]
> gcc9-pr86660.patch
>
> Untested patch. Does this fix it?
I'll try it out.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86660
--- Comment #2 from Jakub Jelinek ---
Created attachment 44437
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44437&action=edit
gcc9-pr86660.patch
Untested patch. Does this fix it?
10 matches
Mail list logo
