linux-kernel - Re: [Patch v3 Part2 3/9] x86/microcode/intel: Fix collect_cpu

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 31 Jan 2023 09:41:09 -0800
From:   Ashok Raj <ashok.raj@...el.com>
To:     "Luck, Tony" <tony.luck@...el.com>
CC:     Borislav Petkov <bp@...en8.de>,
        Thomas Gleixner <tglx@...utronix.de>,
        LKML <linux-kernel@...r.kernel.org>, x86 <x86@...nel.org>,
        Ingo Molnar <mingo@...nel.org>,
        "Hansen, Dave" <dave.hansen@...el.com>,
        "Schofield, Alison" <alison.schofield@...el.com>,
        "Chatre, Reinette" <reinette.chatre@...el.com>,
        Tom Lendacky <thomas.lendacky@....com>,
        "Stefan Talpalaru" <stefantalpalaru@...oo.com>,
        David Woodhouse <dwmw2@...radead.org>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Jonathan Corbet <corbet@....net>,
        "Rafael J . Wysocki" <rafael@...nel.org>,
        Peter Zilstra <peterz@...radead.org>,
        "Lutomirski, Andy" <luto@...nel.org>,
        "andrew.cooper3@...rix.com" <andrew.cooper3@...rix.com>,
        "Ostrovsky, Boris" <boris.ostrovsky@...cle.com>,
        Martin Pohlack <mpohlack@...zon.de>,
        Ashok Raj <ashok.raj@...el.com>
Subject: Re: [Patch v3 Part2 3/9] x86/microcode/intel: Fix collect_cpu_info()
 to reflect current microcode

On Tue, Jan 31, 2023 at 09:34:14AM -0800, Tony Luck wrote:
> >> Currently collect_cpu_info() is only returning what was cached earlier
> >> instead of reading the current revision from the proper MSR.
> >
> > Because this is how this is supposed to work: you collect what's
> > currently applied - which should be exactly the same as what's in the
> > MSR - and then see if there's a new patch in the cache and if so, update
> > it.
> 
> But those get out of step when applying ucode on one logical CPU does
> an update to other(s) (in this case the HT sibling for the same core).
> 
> Would you prefer that Ashok leave collect_cpu_info() as-is, and create
> a new function to read the MSR?
> 

Untested patch below. I think we could add a new function if needed as you
suggest.

Boris, do you have the same dependency to take the siblings through an
update to update some portions of the ucode? or they are only for early
loading?


diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c
index d144f918a896..0038690a5d4b 100644
--- a/arch/x86/kernel/cpu/microcode/amd.c
+++ b/arch/x86/kernel/cpu/microcode/amd.c
@@ -673,8 +673,10 @@ static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig)
 	 * mc_bp_resume() can call apply_microcode()
 	 */
 	p = find_patch(cpu);
-	if (p && (p->patch_id == csig->rev))
+	if (p) {
+		csig->rev = c->microcode = ((struct microcode_amd *)p->data)->hdr.patch_id;
 		uci->mc = p->data;
+	}
 
 	pr_info("CPU%d: patch_level=0x%08x\n", cpu, csig->rev);