-
Notifications
You must be signed in to change notification settings - Fork 546
Add PTX Debug #88
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
base: dev
Are you sure you want to change the base?
Add PTX Debug #88
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Overall I think this looks okay.
However, the code to emit the debug PTX is always run (even if we are not debugging anything) - I think this will only happen when the code is loaded so the perf hit is minimal, but interleaving all the debugging code with the normal execution code makes it really hard to read/modify. Can we try to pull all the code creating the debug code into separate function instead of sprinkling it all over the normal flow?
if( num_elem >= 6 ) out[5] = o.reg6_num(); | ||
if( num_elem >= 7 ) out[6] = o.reg7_num(); | ||
if( num_elem >= 8 ) out[7] = o.reg8_num(); | ||
if( num_elem >= 1 ) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Clearly this works - but can we separate all the debug code creation from the normal code flow?
All this code is actually called even if we not doing the debug part...
Can we pull the creation of the debug ptx into a separate function, only called when debugging is enabled?
* Gcc13 support (gpgpu-sim#87) * Update setup_environment Sims should work with latest CUDA * Update setup_environment Throwing an error on updating CUDA is a bit much. Let's warn them * Update main.yml Turning off the format code for now. Currently, set up to run on an internal cluster. Needs testing on docker. * fix gcc13 unit64 missing header --------- Co-authored-by: Tor Aamodt <[email protected]> Co-authored-by: Tim Rogers <[email protected]> * Cuda12 support (gpgpu-sim#86) * Update setup_environment Sims should work with latest CUDA * Update setup_environment Throwing an error on updating CUDA is a bit much. Let's warn them * Update main.yml Turning off the format code for now. Currently, set up to run on an internal cluster. Needs testing on docker. * added support to cuda 12, by predicating texuture cache * format code --------- Co-authored-by: Tor Aamodt <[email protected]> Co-authored-by: Tim Rogers <[email protected]> * Changed to use the new image * merge upstream (gpgpu-sim#88) * Update setup_environment Sims should work with latest CUDA * Update setup_environment Throwing an error on updating CUDA is a bit much. Let's warn them * Update main.yml Turning off the format code for now. Currently, set up to run on an internal cluster. Needs testing on docker. --------- Co-authored-by: Tor Aamodt <[email protected]> * Updated docker image * Update CMakeLists.txt Support CUDA 12 --------- Co-authored-by: Ahmad Alawneh <[email protected]> Co-authored-by: Tor Aamodt <[email protected]> Co-authored-by: Tim Rogers <[email protected]> Co-authored-by: Ni Kang <[email protected]>
add relevant store after each instruction in ptx code