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-next>] [day] [month] [year] [list]
Date:   Mon, 11 Jan 2021 16:17:03 +0100
From:   Max Krummenacher <max.oss.09@...il.com>
To:     Max Krummenacher <max.oss.09@...il.com>
Cc:     Max Krummenacher <max.krummenacher@...adex.com>,
        Lucas Stach <l.stach@...gutronix.de>,
        linux-arm-kernel@...ts.infradead.org,
        Fabio Estevam <festevam@...il.com>,
        Rouven Czerwinski <r.czerwinski@...gutronix.de>,
        linux-kernel@...r.kernel.org,
        Ahmad Fatoum <a.fatoum@...gutronix.de>,
        Pengutronix Kernel Team <kernel@...gutronix.de>,
        Sascha Hauer <s.hauer@...gutronix.de>,
        Russell King <linux@...linux.org.uk>,
        Shawn Guo <shawnguo@...nel.org>,
        NXP Linux Team <linux-imx@....com>
Subject: [PATCH 0/1] ARM: imx: build suspend-imx6.S with arm instruction set

When the kernel is configured to use the Thumb-2 instruction set
"suspend-to-memory" fails to resume while in ARM mode it works as
expected.
(I used imx_v6_v7_defconfig and deselected ARCH_MULTI_V6 and selected
THUMB2_KERNEL)

The system prints what is expected when suspending but an event of a
wakeup source does give no output.

root@...ibri-imx6ull:~# echo mem > /sys/power/state

[   58.610809] PM: suspend entry (deep)
[   58.629354] Filesystems sync: 0.014 seconds
[   58.653411] Freezing user space processes ... (elapsed 0.001 seconds) done.
[   58.661941] OOM killer disabled.
[   58.665176] Freezing remaining freezable tasks ... (elapsed 0.001 seconds) done.
[   58.674028] printk: Suspending console(s) (use no_console_suspend to debug)

-> trigger wakeup event, no reaction.

It looks like the CPU resumes unconditionally in ARM instruction mode
and then chokes on the presented Thumb-2 code it should execute on resume.
With the following code change resume succeeds.

--- a/arch/arm/mach-imx/suspend-imx6.S
+++ b/arch/arm/mach-imx/suspend-imx6.S
@@ -287,11 +286,20 @@ rbc_loop:
 	bne	rbc_loop

 	/* Zzz, enter stop mode */
 	wfi
+#ifdef CONFIG_THUMB2_KERNEL
+	/* i.MX CPUs seem to leave stop mode set to ARM instruction set */
+	.arm
+#endif
 	nop
 	nop
 	nop
 	nop
+#ifdef CONFIG_THUMB2_KERNEL
+	/* switch to Thumb2 mode */
+	sub     pc, pc, #3
+	.thumb
+#endif

 	/*

I propose however to compile the whole file in ARM mode and have the
linker taking care of the ARM/Thumb-2 switching. This would also keep
the code working if a i.MX CPU variant exists that resumes in the same
mode in which it went to sleep.



Max Krummenacher (1):
  ARM: imx: build suspend-imx6.S with arm instruction set

 arch/arm/mach-imx/suspend-imx6.S | 1 +
 1 file changed, 1 insertion(+)

-- 
2.26.2

Powered by blists - more mailing lists