On 11/4/2022 3:50 AM, Borislav Petkov wrote:
> As to patch 8, that metadata checking should not be part of
> microcode_intel_sanity_check() but a separate function. Along with
> microcode_intel_find_meta_data() - all those should go into the IFS
> thing.
>
> When microcode loading ends up really needing metadata, *then*
> that functionality should be lifted into a more fitting place like
> cpu/intel.c
Thanks for the pointer. I will move the microcode_intel_find_meta_data()
function into IFS driver and also drop the metadata checking from the exported
microcode_intel_sanity_check()
Wanted to check with you, if it is okay to rename the first "reserved" field
in microcode_header_intel to "metasize" today (as shown in the diff below)
or would you prefer to do that too at a later point ? (doing so today will help to
avoid redefining an IFS specific header struct, with this as the only change )
diff --git a/arch/x86/include/asm/microcode_intel.h b/arch/x86/include/asm/microcode_intel.h
index a4d2ed43193c..bec23c11ca52 100644
--- a/arch/x86/include/asm/microcode_intel.h
+++ b/arch/x86/include/asm/microcode_intel.h
@@ -14,7 +14,8 @@ struct microcode_header_intel {
unsigned int pf;
unsigned int datasize;
unsigned int totalsize;
- unsigned int reserved[3];
+ unsigned int metasize;
+ unsigned int reserved[2];
};
Jithu
