From: Arnd Bergmann arnd@arndb.de
commit f3d964673b2f1c5d5c68c77273efcf7103eed03b upstream.
As Stepan Golosunov points out, there is a small mistake in the get_timespec64() function in the kernel. It was originally added under the assumption that CONFIG_64BIT_TIME would get enabled on all 32-bit and 64-bit architectures, but when the conversion was done, it was only turned on for 32-bit ones.
The effect is that the get_timespec64() function never clears the upper half of the tv_nsec field for 32-bit tasks in compat mode. Clearing this is required for POSIX compliant behavior of functions that pass a 'timespec' structure with a 64-bit tv_sec and a 32-bit tv_nsec, plus uninitialized padding.
The easiest fix for linux-5.1 is to just make the Kconfig symbol unconditional, as it was originally intended. As a follow-up, the #ifdef CONFIG_64BIT_TIME can be removed completely..
Note: for native 32-bit mode, no change is needed, this works as designed and user space should never need to clear the upper 32 bits of the tv_nsec field, in or out of the kernel.
Fixes: 00bf25d693e7 ("y2038: use time32 syscall names on 32-bit") Signed-off-by: Arnd Bergmann arnd@arndb.de Signed-off-by: Thomas Gleixner tglx@linutronix.de Cc: Joseph Myers joseph@codesourcery.com Cc: libc-alpha@sourceware.org Cc: linux-api@vger.kernel.org Cc: Deepa Dinamani deepa.kernel@gmail.com Cc: Lukasz Majewski lukma@denx.de Cc: Stepan Golosunov stepan@golosunov.pp.ru Link: https://lore.kernel.org/lkml/20190422090710.bmxdhhankurhafxq@sghpc.golosunov... Link: https://lkml.kernel.org/r/20190429131951.471701-1-arnd@arndb.de Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org
--- arch/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
--- a/arch/Kconfig +++ b/arch/Kconfig @@ -764,7 +764,7 @@ config COMPAT_OLD_SIGACTION bool
config 64BIT_TIME - def_bool ARCH_HAS_64BIT_TIME + def_bool y help This should be selected by all architectures that need to support new system calls with a 64-bit time_t. This is relevant on all 32-bit