Fix a CI failure for any distro using git >= 2.35.3

[superm1]

Type of pull request:

• New plugin (Please include new plugin checklist)
• Code fix
• Feature
• Documentation

[superm1]

I saw you had this branch, which it seems like the right move to me if it fixes Fedora CI.

[hughsie]

which it seems like the right move to me if it fixes Fedora CI

I think if is the key word there. I had high hopes; but I regen'd the containers with this change and Fedora (or any distro with the newest git) still fails. I still need to test just doing the v2 to v3 change, but gah.

[hughsie]

Also see this: actions/checkout#770 -- it's all a bit of a mess.

[hughsie]

I think merging this is probably the correct thing to do, although I'm still not sure what the root issue is. I suspect meson is calling git and getting the permission error caused by the new git feature, and swallowing it, and returning as if there's no git tree there.

[hughsie]

@superm1 getting there! I'm now getting what I think is a false positive:

In function ‘fu_dell_execute_simple_smi’,
    inlined from ‘fu_dell_detect_dock’ at ../plugins/dell/fu-dell-smi.c:106:7:
../plugins/dell/fu-dell-smi.c:84:13: error: ‘dell_simple_ci_smi’ reading 16 bytes from a region of size 4 [-Werror=stringop-overread]
   84 |         if (dell_simple_ci_smi(class, select, obj->input, obj->output)) {
      |             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../plugins/dell/fu-dell-smi.c:84:13: note: referencing argument 3 of type ‘const u32[4]’
../plugins/dell/fu-dell-smi.c:84:13: note: referencing argument 4 of type ‘u32[4]’
In file included from ../plugins/dell/fu-dell-smi.h:12,
		 from ../plugins/dell/fu-dell-smi.c:9:
/usr/include/smbios_c/smi.h: In function ‘fu_dell_detect_dock’:
/usr/include/smbios_c/smi.h:40:26: note: in a call to function ‘dell_simple_ci_smi’
   40 | LIBSMBIOS_C_DLL_SPEC int dell_simple_ci_smi(u16 smiClass, u16 select, const u32 args[4], u32 res[4]);
      |                          ^~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors

Sanity check please?

[superm1]

@superm1 getting there! I'm now getting what I think is a false positive:

In function ‘fu_dell_execute_simple_smi’,
    inlined from ‘fu_dell_detect_dock’ at ../plugins/dell/fu-dell-smi.c:106:7:
../plugins/dell/fu-dell-smi.c:84:13: error: ‘dell_simple_ci_smi’ reading 16 bytes from a region of size 4 [-Werror=stringop-overread]
   84 |         if (dell_simple_ci_smi(class, select, obj->input, obj->output)) {
      |             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../plugins/dell/fu-dell-smi.c:84:13: note: referencing argument 3 of type ‘const u32[4]’
../plugins/dell/fu-dell-smi.c:84:13: note: referencing argument 4 of type ‘u32[4]’
In file included from ../plugins/dell/fu-dell-smi.h:12,
		 from ../plugins/dell/fu-dell-smi.c:9:
/usr/include/smbios_c/smi.h: In function ‘fu_dell_detect_dock’:
/usr/include/smbios_c/smi.h:40:26: note: in a call to function ‘dell_simple_ci_smi’
   40 | LIBSMBIOS_C_DLL_SPEC int dell_simple_ci_smi(u16 smiClass, u16 select, const u32 args[4], u32 res[4]);
      |                          ^~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors

Sanity check please?

Yeah looks false positive to me too. I wonder if you can cast an array to make it happier. But if not then a pragma to ignore makes sense to me