cuda: Update SUBDBG format specifier from %lu to %u to avoid compilation warning

[Treece-Burgess]

Pull Request Descriptionl

When configuring PAPI as follows ./configure --prefix=$PWD/test-install --with-components="cuda" --with-debug=yes and subsequently building on Heimdall at Oregon (1 * NVIDIA GeForce RTX 5080) you will run into the below compilation warnings:

components/cuda/cupti_event_and_metric.c:1591:12: warning: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 3 has type ‘uint32_t’ {aka ‘unsigned int’} [-Wformat=]
 1591 |     SUBDBG("ENTER: event_code: %lu, modifier: %d\n", *event_code, modifier);
      |            ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  ~~~~~~~~~~~
      |                                                      |
      |                                                      uint32_t {aka unsigned int}
./papi_debug.h:60:115: note: in definition of macro ‘PAPIDEBUG’
   60 | #define PAPIDEBUG(level,format, args...) { if(_papi_hwi_debug&level){DEBUGLABEL(DEBUGLEVEL(level));fprintf(stderr,format, ## args);}}
      |                                                                                                                   ^~~~~~
components/cuda/cupti_event_and_metric.c:1591:5: note: in expansion of macro ‘SUBDBG’
 1591 |     SUBDBG("ENTER: event_code: %lu, modifier: %d\n", *event_code, modifier);
      |     ^~~~~~
components/cuda/cupti_event_and_metric.c:1591:34: note: format string is defined here
 1591 |     SUBDBG("ENTER: event_code: %lu, modifier: %d\n", *event_code, modifier);
      |                                ~~^
      |                                  |
      |                                  long unsigned int
      |                                %u
+ test  != 

This PR resolves this compilation warning and closes Issue #519 by updating a format specifier from %lu to %u.

Testing

Setup

This PR was tested on Heimdall at Oregon with:

• CPU: AMD Ryzen 9 9950X 16-Core Processor
• GPU: 1 * NVIDIA GeForce RTX 5080
• OS: Ubuntu 24.04.3
• Cuda Toolkit: 12.9

Results

• Compilation and PAPI build: ✅ (compilation warnings did not occur)
• PAPI utilities*: ✅
• Cuda component tests: ✅ (Note: Running runtest.sh will result in test failures as the event cuda:::dram__bytes_read:stat=sum:device=0 does not exist on Heimdall)

* - papi_component_avail, papi_native_avail, papi_command_line

Author Checklist

• Description
  Why this PR exists. Reference all relevant information, including background, issues, test failures, etc
• Commits
  Commits are self contained and only do one thing
  Commits have a header of the form: module: short description
  Commits have a body (whenever relevant) containing a detailed description of the addressed problem and its solution
• Tests
  The PR needs to pass all the tests

[dbarry9]

I tested these changes on a system containing the NVIDIA Blackwell architecture and using CUDA 12.9.0 and GCC 13.3.0, and the compiler warning from Issue #519 no longer occurs. The PAPI utilities (papi_avail, papi_native_avail, papi_component_avail, papi_hardware_avail, and papi_mem_info) and cuda component tests behave as expected.

@Treece-Burgess Can you rebase so that we can merge this PR?