linux-kernel - Re: kernel BUG at arch/arm/kvm/coproc.c:1406!

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] [day] [month] [year] [list]

Message-ID: <3d719d5f-d7b9-9bba-4fa8-61175a9b05a9@arm.com>
Date:   Thu, 4 Oct 2018 09:19:45 +0100
From:   Marc Zyngier <marc.zyngier@....com>
To:     Florian Fainelli <f.fainelli@...il.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        kvmarm@...ts.cs.columbia.edu,
        linux ARM <linux-arm-kernel@...ts.infradead.org>,
        Russell King <linux@...linux.org.uk>, christoffer.dall@....com
Subject: Re: kernel BUG at arch/arm/kvm/coproc.c:1406!

Hi Florian,

On 04/10/18 04:02, Florian Fainelli wrote:
> Hi,
> 
> Seeing the following BUG_ON() being triggered on a Lamobo R1
> (SUN7I/Cortex-A7), did not have the time to run a bisection since I was
> chasing another regression. My guess would be that
> 3e8a8a50c7ef1a4f71921731f0e1748a7f70ddaa ("KVM: arm: vgic-v3: Add
> support for ICC_SGI0R and ICC_ASGI1R accesses") is the culprit since it
> does add new entries to cp15_regs[].

OK, this is quite embarrassing. I remember fixing that exact bug before
the 4.19 merge window, and was 100% convinced that I had merged the fix.

Turns out I didn't. 

Can you verify that the following patch works for you?

diff --git a/arch/arm/kvm/coproc.c b/arch/arm/kvm/coproc.c
index 450c7a4fbc8a..cb094e55dc5f 100644
--- a/arch/arm/kvm/coproc.c
+++ b/arch/arm/kvm/coproc.c
@@ -478,15 +478,15 @@ static const struct coproc_reg cp15_regs[] = {
 
 	/* ICC_SGI1R */
 	{ CRm64(12), Op1( 0), is64, access_gic_sgi},
-	/* ICC_ASGI1R */
-	{ CRm64(12), Op1( 1), is64, access_gic_sgi},
-	/* ICC_SGI0R */
-	{ CRm64(12), Op1( 2), is64, access_gic_sgi},
 
 	/* VBAR: swapped by interrupt.S. */
 	{ CRn(12), CRm( 0), Op1( 0), Op2( 0), is32,
 			NULL, reset_val, c12_VBAR, 0x00000000 },
 
+	/* ICC_ASGI1R */
+	{ CRm64(12), Op1( 1), is64, access_gic_sgi},
+	/* ICC_SGI0R */
+	{ CRm64(12), Op1( 2), is64, access_gic_sgi},
 	/* ICC_SRE */
 	{ CRn(12), CRm(12), Op1( 0), Op2(5), is32, access_gic_sre },
 

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...