From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
 id 56BEE3858C27; Mon, 25 Oct 2021 11:35:14 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 56BEE3858C27
From: "ardb at kernel dot org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/102926] New: TLS register value is spilled to the stack
 instead of reloaded from the system register
Date: Mon, 25 Oct 2021 11:35:14 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 12.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ardb at kernel dot org
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status
 bug_severity priority component assigned_to reporter target_milestone
 cf_gcctarget
Message-ID: <bug-102926-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-BeenThere: gcc-bugs@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-bugs mailing list <gcc-bugs.gcc.gnu.org>
List-Unsubscribe: <https://gcc.gnu.org/mailman/options/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-request@gcc.gnu.org?subject=help>
List-Subscribe: <https://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Mon, 25 Oct 2021 11:35:14 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D102926

            Bug ID: 102926
           Summary: TLS register value is spilled to the stack instead of
                    reloaded from the system register
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ardb at kernel dot org
  Target Milestone: ---
            Target: ARM

The code below uses the hardware TLS register on ARM, and instead of reload=
ing
its value directly from the system register, it spills its value to the sta=
ck.
This is suboptimal, and given that the TLS register is being proposed as an
alternative reference for the stack protector canary, it is also a security
concern, as an attacker that controls the stack may be able to control both
sides of the equation in the stack protector check occurring at the end of =
the
function.

Instead, I would expect any subsequent uses of the thread pointer to simply
issue the MRC again, which doesn't touch memory.

$ cat /tmp/spill.c=20

    int foo(int);
    int bar(void)
    {
        int *l =3D __builtin_thread_pointer();

        return foo(l[0]) + l[1];
    }

$ arm-linux-gnueabihf-gcc -o - -S /tmp/spill.c -O3 -mtp=3Dcp15 -ffixed-r4
-ffixed-r5 -ffixed-r6 -ffixed-r7 -ffixed-r8 -ffixed-r9 -ffixed-r10
-fno-omit-frame-pointer
        .cpu arm10tdmi
        .arch armv5t
        .fpu softvfp
        .eabi_attribute 20, 1
        .eabi_attribute 21, 1
        .eabi_attribute 23, 3
        .eabi_attribute 24, 1
        .eabi_attribute 25, 1
        .eabi_attribute 26, 2
        .eabi_attribute 30, 2
        .eabi_attribute 34, 0
        .eabi_attribute 18, 4
        .file   "spill.c"
        .text
        .align  2
        .global bar
        .syntax unified
        .arm
        .type   bar, %function
bar:
        @ args =3D 0, pretend =3D 0, frame =3D 8
        @ frame_needed =3D 1, uses_anonymous_args =3D 0
        push    {fp, lr}
        add     fp, sp, #4
        sub     sp, sp, #8
        mrc     p15, 0, r3, c13, c0, 3  @ load_tp_hard
        ldr     r0, [r3]
        str     r3, [fp, #-8]
        bl      foo
        ldr     r3, [fp, #-8]
        ldr     r3, [r3, #4]
        add     r0, r0, r3
        sub     sp, fp, #4
        @ sp needed
        pop     {fp, pc}
        .size   bar, .-bar
        .ident  "GCC: (GNU) 12.0.0 20211024 (experimental)"
        .section        .note.GNU-stack,"",%progbits=