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  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Sat, 23 Mar 2019 09:59:34 +0100
From:   Borislav Petkov <bp@...en8.de>
To:     Pu Wen <puwen@...on.cn>
Cc:     "tglx@...utronix.de" <tglx@...utronix.de>,
        "mingo@...hat.com" <mingo@...hat.com>,
        "hpa@...or.com" <hpa@...or.com>, "x86@...nel.org" <x86@...nel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [RFC PATCH] x86/cpu/hygon: Fix phys_proc_id calculation logic
 for multi-die processor

On Sat, Mar 23, 2019 at 10:13:39AM +0800, Pu Wen wrote:
> Current physical id is computed via "phys_proc_id = initial_apicid >>
> bits". 
> 
> For 4-Die 2 socket system, the physical id of socket 2 is:
> initial_apicid >> bits = 0b1xxxxxx >> 6 = 1.
> The result is true.
> 
> But for 2-Die 2 socket system, the physical id of socket 2 is:
> initial_apicid >> bits = 0b10xxxxx >> 5 = 2,
> and for 1-Die 2 socket system, the physical id of socket 2 is:
> initial_apicid >> bits = 0b100xxxx >> 4 = 4.
> The results are not correct any more.
> 
> So the adjustment for the 1-Die/2-Die 2 socket system is needed.
> And just use ApicId[6], which already defined the right thing, as the
> socket ID.

I understand all that. But let me repeat my question:

So why do you need to do something different than what AMD does?

You said you're programming the initial APIC ID the same as AMD. So why
doesn't this need to be changed in AMD code too but only for hygon?

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

Powered by blists - more mailing lists