If DEBUG_LL and earlyprintk are enabled, and omap-serial.c is compiled
as a module, the kernel boot hangs early as the clocks for serial port
are cut while earlyprintk still uses the port.

The problem is a race between the late_initcall for omap_device (which
idles devices that have no drivers) and the late_initcall in
kernel/printk.c which turns off the earlyconsole.   Any printks
that happen between this omap_device late initcall and the earlyconsole
late initcall will crash when accessing the UART.

The fix is to ensure the omap_device initcall happens after the
earlyconsole initcall.

Reported-by: Tony Lindgren <[email protected]>
Signed-off-by: Kevin Hilman <[email protected]>
---
Based on v3.9-rc8

 arch/arm/mach-omap2/omap_device.c | 2 +-
 arch/arm/mach-omap2/soc.h         | 1 +
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/mach-omap2/omap_device.c 
b/arch/arm/mach-omap2/omap_device.c
index 381be7a..2d20d69 100644
--- a/arch/arm/mach-omap2/omap_device.c
+++ b/arch/arm/mach-omap2/omap_device.c
@@ -879,4 +879,4 @@ static int __init omap_device_late_init(void)
        bus_for_each_dev(&platform_bus_type, NULL, NULL, omap_device_late_idle);
        return 0;
 }
-omap_late_initcall(omap_device_late_init);
+omap_late_initcall_sync(omap_device_late_init);
diff --git a/arch/arm/mach-omap2/soc.h b/arch/arm/mach-omap2/soc.h
index c62116b..de88611 100644
--- a/arch/arm/mach-omap2/soc.h
+++ b/arch/arm/mach-omap2/soc.h
@@ -494,6 +494,7 @@ level(__##fn);
 #define omap_subsys_initcall(fn)       omap_initcall(subsys_initcall, fn)
 #define omap_device_initcall(fn)       omap_initcall(device_initcall, fn)
 #define omap_late_initcall(fn)         omap_initcall(late_initcall, fn)
+#define omap_late_initcall_sync(fn)    omap_initcall(late_initcall_sync, fn)
 
 #endif /* __ASSEMBLY__ */
 
-- 
1.8.2

--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to