linux-kernel - Re: linux-next: Tree for Oct 31 (vboxguest)

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:   Wed, 31 Oct 2018 17:50:47 +0100
From:   Hans de Goede <hdegoede@...hat.com>
To:     Randy Dunlap <rdunlap@...radead.org>,
        Stephen Rothwell <sfr@...b.auug.org.au>,
        Linux-Next Mailing List <linux-next@...r.kernel.org>
Cc:     Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Arnd Bergmann <arnd@...db.de>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: linux-next: Tree for Oct 31 (vboxguest)

Hi,

On 31-10-18 16:51, Randy Dunlap wrote:
> On 10/30/18 8:59 PM, Stephen Rothwell wrote:
>> Hi all,
>>
>> Please do not add any v4.21/v5.1 code to your linux-next included trees
>> until after the merge window closes.
>>
>> Changes since 20181030:
>>
> 
> 
> on i386:
> 
> ld: drivers/virt/vboxguest/vboxguest_core.o: in function `vbg_ioctl_hgcm_call':
> vboxguest_core.c:(.text+0x212b): undefined reference to `vbg_hgcm_call32'

Are you perhaps using weird compiler options?

This call should be optimized out on i386 (and we rely on calls being
removed be dead code elimination to avoid #ifdefs in various places,
iow this is a normal thing to rely on) :

First we have:

	bool f32bit = false;

...

         switch (req_no_size) {
#ifdef CONFIG_COMPAT
         case VBG_IOCTL_HGCM_CALL_32(0):
                 f32bit = true;
                 /* Fall through */
#endif
         case VBG_IOCTL_HGCM_CALL(0):
                 return vbg_ioctl_hgcm_call(gdev, session, f32bit, data);
	}

And then we also have:

static int vbg_ioctl_hgcm_call(struct vbg_dev *gdev,
                                struct vbg_session *session, bool f32bit,
                                struct vbg_ioctl_hgcm_call *call)
{
	...

         if (f32bit)
                 ret = vbg_hgcm_call32(gdev, client_id,
                                       call->function, call->timeout_ms,
                                       VBG_IOCTL_HGCM_CALL_PARMS32(call),
                                       call->parm_count, &call->hdr.rc);
         else
                 ret = vbg_hgcm_call(gdev, client_id,
                                     call->function, call->timeout_ms,
                                     VBG_IOCTL_HGCM_CALL_PARMS(call),
                                     call->parm_count, &call->hdr.rc);
}

So on i386 CONFIG_COMPAT is never set, this f32bit is a 0 const
and the compiler removes the if branch of the if ... else ...

This has been upstream like this since 4.16 without any problems sofar.

Regards,

Hans