linux-kernel - Re: [PATCH] x86/fpu: Remove dynamic features from xcomp_bv for init

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:   Thu, 13 Oct 2022 09:23:20 -0700
From:   "Chang S. Bae" <chang.seok.bae@...el.com>
To:     "Yao, Yuan" <yuan.yao@...el.com>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
CC:     "x86@...nel.org" <x86@...nel.org>,
        "Hansen, Dave" <dave.hansen@...el.com>,
        Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH] x86/fpu: Remove dynamic features from xcomp_bv for
 init_fpstate

On 10/12/2022 8:35 PM, Yao, Yuan wrote:
> 
> The reason is __copy_xstate_to_uabi_buf() copies data from &init_fpstate when the component
> is not existed in the source kernel fpstate (here is the AMX tile component), but the
> AMX TILE bit is removed from init_fpstate due to this patch, so the WARN is triggered and return
> NULL which causes kernel NULL pointer dereference later.

We have this in __copy_xstate_to_uabi_buf() [1]:

	mask = fpstate->user_xfeatures;

	for_each_extended_xfeature(i, mask) {
	...
	}

And the KVM code seems to set dynamic features regardless of the buffer 
reallocation [2]:

	vcpu->arch.guest_fpu.fpstate->user_xfeatures =
		vcpu->arch.guest_supported_xcr0 | XFEATURE_MASK_FPSSE;

The kernel code seems to be aware of this as fpstate_realloc() does [3]:

	if (!guest_fpu)
		newfps->user_xfeatures = curfps->user_xfeatures | xfeatures;

But it updates the 'xfeature' bitmask for all:

	newfps->xfeatures = curfps->xfeatures | xfeatures;

So, I think we can do something like this here:

diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c
index c8340156bfd2..8ea7d0e95f1a 100644
--- a/arch/x86/kernel/fpu/xstate.c
+++ b/arch/x86/kernel/fpu/xstate.c
@@ -1127,8 +1127,12 @@ void __copy_xstate_to_uabi_buf(struct membuf to, 
struct fpstate *fpstate,
          * non-compacted format disabled features still occupy state space,
          * but there is no state to copy from in the compacted
          * init_fpstate. The gap tracking will zero these states.
+        *
+        * In the case of guest fpstate, this user_xfeatures does not
+        * dynamically reflect the capacity of the XSAVE buffer but
+        * xfeatures does. So AND them together.
          */
-       mask = fpstate->user_xfeatures;
+       mask = fpstate->user_xfeatures & fpstate->xfeatures;

Let me also test this by running KVM.

Thanks,
Chang

[1] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/kernel/fpu/xstate.c#n1131
[2] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/kvm/cpuid.c#n346
[3] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/x86/kernel/fpu/xstate.c#n1448