Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Brian Gerst <brgerst@gmail.com>
To:     linux-kernel@vger.kernel.org, x86@kernel.org
Cc:     Thomas Gleixner <tglx@linutronix.de>,
        Borislav Petkov <bp@alien8.de>,
        "H . Peter Anvin" <hpa@zytor.com>,
        Andy Lutomirski <luto@kernel.org>,
        =?UTF-8?q?Mika=20Penttil=C3=A4?= <mpenttil@redhat.com>,
        Brian Gerst <brgerst@gmail.com>
Subject: [PATCH v2 1/6] x86/entry/64: Remove obsolete comment on tracing vs. SYSRET
Date:   Fri, 21 Jul 2023 12:10:12 -0400
Message-ID: <20230721161018.50214-2-brgerst@gmail.com>
In-Reply-To: <20230721161018.50214-1-brgerst@gmail.com>
References: <20230721161018.50214-1-brgerst@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

This comment comes from a time when the kernel attempted to use SYSRET
on all returns to userspace, including interrupts and exceptions.  Ever
since commit fffbb5dc ("Move opportunistic sysret code to syscall code
path"), SYSRET is only used for returning from system calls. The
specific tracing issue listed in this comment is not possible anymore.

Signed-off-by: Brian Gerst <brgerst@gmail.com>
---
 arch/x86/entry/entry_64.S | 19 +++----------------
 1 file changed, 3 insertions(+), 16 deletions(-)

diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
index 91f6818884fa..c01776a51545 100644
--- a/arch/x86/entry/entry_64.S
+++ b/arch/x86/entry/entry_64.S
@@ -166,22 +166,9 @@ SYM_INNER_LABEL(entry_SYSCALL_64_after_hwframe, SYM_L_GLOBAL)
 	jne	swapgs_restore_regs_and_return_to_usermode
 
 	/*
-	 * SYSCALL clears RF when it saves RFLAGS in R11 and SYSRET cannot
-	 * restore RF properly. If the slowpath sets it for whatever reason, we
-	 * need to restore it correctly.
-	 *
-	 * SYSRET can restore TF, but unlike IRET, restoring TF results in a
-	 * trap from userspace immediately after SYSRET.  This would cause an
-	 * infinite loop whenever #DB happens with register state that satisfies
-	 * the opportunistic SYSRET conditions.  For example, single-stepping
-	 * this user code:
-	 *
-	 *           movq	$stuck_here, %rcx
-	 *           pushfq
-	 *           popq %r11
-	 *   stuck_here:
-	 *
-	 * would never get past 'stuck_here'.
+	 * SYSRET cannot restore RF.  It can restore TF, but unlike IRET,
+	 * restoring TF results in a trap from userspace immediately after
+	 * SYSRET.
 	 */
 	testq	$(X86_EFLAGS_RF|X86_EFLAGS_TF), %r11
 	jnz	swapgs_restore_regs_and_return_to_usermode
-- 
2.41.0