linux-kernel - Re: [PATCH v3 01/13] x86/retpoline: Add initial retpoline support

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Mon, 8 Jan 2018 07:55:20 +0000
From:   "Woodhouse, David" <dwmw@...zon.co.uk>
To:     Josh Poimboeuf <jpoimboe@...hat.com>, Borislav Petkov <bp@...e.de>
CC:     "tglx@...utronix.de" <tglx@...utronix.de>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "tim.c.chen@...ux.intel.com" <tim.c.chen@...ux.intel.com>,
        "peterz@...radead.org" <peterz@...radead.org>,
        "torvalds@...ux-foundation.org" <torvalds@...ux-foundation.org>,
        "ak@...ux.intel.com" <ak@...ux.intel.com>,
        "riel@...hat.com" <riel@...hat.com>,
        "keescook@...gle.com" <keescook@...gle.com>,
        "gnomes@...rguk.ukuu.org.uk" <gnomes@...rguk.ukuu.org.uk>,
        "pjt@...gle.com" <pjt@...gle.com>,
        "dave.hansen@...el.com" <dave.hansen@...el.com>,
        "luto@...capital.net" <luto@...capital.net>,
        "jikos@...nel.org" <jikos@...nel.org>,
        "gregkh@...ux-foundation.org" <gregkh@...ux-foundation.org>
Subject: Re: [PATCH v3 01/13] x86/retpoline: Add initial retpoline support

On Sun, 2018-01-07 at 23:06 -0600, Josh Poimboeuf wrote:
> 
> Here's the use case I had in mind before.  With paravirt,
> 
>   ENABLE_INTERRUPTS(CLBR_NONE)
> 
> becomes
> 
>   push  %rax
>   call  *pv_irq_ops.irq_enable
>   pop   %rax
> 
> and I wanted to apply those instructions with an alternative.  It
> doesn't work currently because the 'call' isn't first.

I believe Borislav has made that work... however, if you're literally
doing the above then you'd be introducing new indirect branches which
is precisely what I've been trying to eliminate.

I believe I was told to stop prodding at pvops and just trust that they
all get turned into *direct* jumps at runtime. For example the above
call would not be literally 'call *pv_irq_ops.irq_enable', because by
the time the pvops are patched we'd *know* the final value of the
irq_enable method, and we'd turn it into a *direct* call instead.

Do I need to start looking at pvops again?
Download attachment "smime.p7s" of type "application/x-pkcs7-signature" (5210 bytes)