Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ocelot/translator/implementation/PTXToLLVMTranslator.cpp:2686: void translator::PTXToLLVMTranslator::_translateCvta(const ir::PTXInstruction&): Assertion `false' failed. #83

Open
jwang323 opened this issue Mar 29, 2015 · 1 comment

Comments

@jwang323
Copy link
Contributor

From volyrkr on June 04, 2013 20:20:05

What steps will reproduce the problem? 1. Run CUDA program linked with gpuocelot as in documentation. What is the expected output? What do you see instead? - Should run normally with "llvm" backend.
Runs normally (but slow) only with "emulated" backend. What version of the product are you using? On what operating system? -Latest from svn, latest llvm-3.4svn, Gentoo GNU/Linux linux-3.8.8, glibc-2.15, gcc-4.6.3 Please provide any additional information below. -Could not find any clues on google.

Original issue: http://code.google.com/p/gpuocelot/issues/detail?id=84

@jwang323
Copy link
Contributor Author

From volyrkr on June 04, 2013 19:18:06

I don't have time to learn all the details so I found the type that was 'invalid' - ir::PTXInstruction::Const and added it to the choice list.
Brute force worked.
Why Const was not included?
Thanks.

Index: PTXToLLVMTranslator.cpp

--- PTXToLLVMTranslator.cpp ( revision 2235 )
+++ PTXToLLVMTranslator.cpp (working copy)
@@ -2665,6 +2665,7 @@
{
case ir::PTXInstruction::Shared: /* fall through */
case ir::PTXInstruction::Local:

  •   case ir::PTXInstruction::Const:
    {
        assertM( i.addressSpace != ir::PTXInstruction::Local
            || i.a.addressMode != ir::PTXOperand::Address
    

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant