|SUCCESS| power/amd_uncore: detect esmi lib64 using pkgconfig

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: SUCCESS
https://dpdk.org/patch/162785

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-03-31

git sha: 20260331040958.126869-1-sivaprasad.tummala at amd.com
Author: Sivaprasad Tummala <sivaprasad.tummala at amd.com>
Subject: [PATCH] power/amd_uncore: detect esmi lib64 using pkgconfig

This commit replaces manual library detection with pkg-config-based
dependency resolution for e_smi_lib64. The change addresses potential
incompatibility issues with find_library() by using .pc files.

> diff --git a/drivers/power/amd_uncore/meson.build b/drivers/power/amd_uncore/meson.build
> index cab260b64e..378eb72e50 100644
> --- a/drivers/power/amd_uncore/meson.build
> +++ b/drivers/power/amd_uncore/meson.build
> @@ -7,14 +7,12 @@ if not is_linux
>      subdir_done()
>  endif
>  
> -ESMI_header = '#include<e_smi/e_smi.h>'
> -lib = cc.find_library('e_smi64', required: false)
> -if not lib.found() or not cc.links(min_c_code, dependencies: lib)
> +dep = dependency('e_smi_lib64', required: false, method: 'pkg-config')
> +if not dep.found()

Does this code remove the header availability check? The original code
appears to define ESMI_header, and though it's not shown in the
previous context being removed, does the check for the header's
presence get lost? If the header check happened via
cc.links(min_c_code, dependencies: lib), replacing it with just
dep.found() might allow builds to proceed without verifying that the
header file is actually available.

>      build = false
> -    reason = 'missing dependency, "libe_smi"'
> -else
> -    ext_deps += lib
> +    reason = 'missing dependency, "e_smi_lib64"'
> +    subdir_done()
>  endif
> -
>  sources = files('amd_uncore.c')
> +ext_deps += dep
>  deps += ['power']