lmsensors component: Replace fprintf with SUBDBG

[Treece-Burgess]

Pull Request Description

Currently in the lmsensors component there are two fprintf calls. This can lead to error messages printing even when you do not set PAPI_DEBUG. As an example, on an AMD Epyc 7413 with lmsensors 3.4 the following will occur:

[tburgess@gilgamesh bin]$ ./papi_command_line lmsensors:::amdgpu-pci-e300.mem.temp3_crit

This utility lets you add events from the command line interface to see if they work.

libsensors(): Could not read event #26!
.
.
.
libsensors(): Could not read event #75!
Successfully added: lmsensors:::amdgpu-pci-e300.mem.temp3_crit

lmsensors:::amdgpu-pci-e300.mem.temp3_crit : 	94000 

This PR resolves this issue by replacing the fprintf's with SUBDBG's (Tested on an AMD Epyc 7413 with lmsensors 3.4).

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 am testing this pull request.

[dbarry9]

On Gilgamesh using GCC 12.2.0, I get the following compiler warnings:

components/lmsensors/linux-lmsensors.c: In function 'createNativeEvents':
components/lmsensors/linux-lmsensors.c:204:18: warning: passing argument 2 of 'fprintf' from incompatible pointer type [-Wincompatible-pointer-types]
  204 |                  SUBDBG( stderr, "ERROR: Can't get label of feature %s!\n", feature->name );
      |                  ^~~~~~
      |                  |
      |                  FILE *
In file included from /home/users/dbarry/apps/lmsensors/include/sensors/sensors.h:25,
                 from components/lmsensors/linux-lmsensors.c:32:
/usr/include/stdio.h:327:44: note: expected 'const char * restrict' but argument is of type 'FILE *'
  327 |                     const char *__restrict __format, ...);
      |                     ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~
components/lmsensors/linux-lmsensors.c: In function 'getEventValue':
components/lmsensors/linux-lmsensors.c:262:17: warning: passing argument 2 of 'fprintf' from incompatible pointer type [-Wincompatible-pointer-types]
  262 |                 SUBDBG( stderr, "libsensors(): Could not read event #%d!\n", event_id );
      |                 ^~~~~~
      |                 |
      |                 FILE *
/usr/include/stdio.h:327:44: note: expected 'const char * restrict' but argument is of type 'FILE *'
  327 |                     const char *__restrict __format, ...);
      |                     ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~

I believe this can remedied by removing the stderr arguments.

[Treece-Burgess]

On Gilgamesh using GCC 12.2.0, I get the following compiler warnings:

components/lmsensors/linux-lmsensors.c: In function 'createNativeEvents':
components/lmsensors/linux-lmsensors.c:204:18: warning: passing argument 2 of 'fprintf' from incompatible pointer type [-Wincompatible-pointer-types]
  204 |                  SUBDBG( stderr, "ERROR: Can't get label of feature %s!\n", feature->name );
      |                  ^~~~~~
      |                  |
      |                  FILE *
In file included from /home/users/dbarry/apps/lmsensors/include/sensors/sensors.h:25,
                 from components/lmsensors/linux-lmsensors.c:32:
/usr/include/stdio.h:327:44: note: expected 'const char * restrict' but argument is of type 'FILE *'
  327 |                     const char *__restrict __format, ...);
      |                     ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~
components/lmsensors/linux-lmsensors.c: In function 'getEventValue':
components/lmsensors/linux-lmsensors.c:262:17: warning: passing argument 2 of 'fprintf' from incompatible pointer type [-Wincompatible-pointer-types]
  262 |                 SUBDBG( stderr, "libsensors(): Could not read event #%d!\n", event_id );
      |                 ^~~~~~
      |                 |
      |                 FILE *
/usr/include/stdio.h:327:44: note: expected 'const char * restrict' but argument is of type 'FILE *'
  327 |                     const char *__restrict __format, ...);
      |                     ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~

I believe this can remedied by removing the stderr arguments.

@dbarry9 Thank you for catching that, yes that is correct. I have updated the PR and tested on Gilgamesh myself with GCC 12.2.0 and do not notice the compiler warning now. Please let me know if it is the same for yourself.

[dbarry9]

@Treece-Burgess That resolved the compiler warning on my end, too.

I have now tested this pull request on Gilgamesh (AMD Zen3 + NVIDIA H100 + AMD MI250X) and Frontier (AMD Zen3 + MI250X). The utilities work as expected, and the lmsensors component tests' output is the same as in the master branch. Although, as an unrelatd issue, the lmsensors_read component test currently produces an error in the master branch and the branch in this PR:

Component index for lmsensors: 2
PAPI Error: Error Code -17,Component Index isn't set
FAILED!!!
Line # 95 Error in PAPI_start: Component Index isn't set
Some tests require special hardware, permissions, OS, compilers
or library versions. PAPI may still function perfectly on your
system without the particular feature being tested here.

[Treece-Burgess]

@Treece-Burgess That resolved the compiler warning on my end, too.

I have now tested this pull request on Gilgamesh (AMD Zen3 + NVIDIA H100 + AMD MI250X) and Frontier (AMD Zen3 + MI250X). The utilities work as expected, and the lmsensors component tests' output is the same as in the master branch. Although, as an unrelatd issue, the lmsensors_read component test currently produces an error in the master branch and the branch in this PR:

Component index for lmsensors: 2
PAPI Error: Error Code -17,Component Index isn't set
FAILED!!!
Line # 95 Error in PAPI_start: Component Index isn't set
Some tests require special hardware, permissions, OS, compilers
or library versions. PAPI may still function perfectly on your
system without the particular feature being tested here.

I will take a look at the lmsensors_read test to see why it is failing.

[Treece-Burgess]

@Treece-Burgess That resolved the compiler warning on my end, too.

I have now tested this pull request on Gilgamesh (AMD Zen3 + NVIDIA H100 + AMD MI250X) and Frontier (AMD Zen3 + MI250X). The utilities work as expected, and the lmsensors component tests' output is the same as in the master branch. Although, as an unrelatd issue, the lmsensors_read component test currently produces an error in the master branch and the branch in this PR:

Component index for lmsensors: 2
PAPI Error: Error Code -17,Component Index isn't set
FAILED!!!
Line # 95 Error in PAPI_start: Component Index isn't set
Some tests require special hardware, permissions, OS, compilers
or library versions. PAPI may still function perfectly on your
system without the particular feature being tested here.

For documentation purposes, I will be creating a separate PR to handle the error of the lmsensors_read component test.