From mboxrd@z Thu Jan 1 00:00:00 1970 From: moon.linux@yahoo.com (Anand Moon) Date: Fri, 5 Dec 2014 20:21:06 +0530 Subject: [PATCH] do_work_pending: Enable interrupts in do_work_pending() Message-ID: <1417791066-5458-1-git-send-email-moon.linux@yahoo.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org A recent change triggered a WARN_ON that interrupts were disabled, and in fact other architectures enable interrupts uniformly for their related do_work_pending() type code, and all the things we call from this routine appear to expect interrupts to be enabled. So, enable them. Below is the warning. [ 12.094106] systemd-udevd[1643]: starting version 204 [ 13.466537] ------------[ cut here ]------------ [ 13.469750] WARNING: CPU: 1 PID: 1645 at kernel/locking/lockdep.c:3536 check_flags+0x7c/0x1d0() [ 13.478397] DEBUG_LOCKS_WARN_ON(current->hardirqs_enabled) [ 13.483663] Modules linked in: [ 13.486863] CPU: 1 PID: 1645 Comm: systemd-udevd Not tainted 3.17.4-arm7 #6 [ 13.493851] [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [ 13.501563] [] (show_stack) from [] (dump_stack+0x80/0xac) [ 13.508753] [] (dump_stack) from [] (warn_slowpath_common+0x60/0x84) [ 13.516808] [] (warn_slowpath_common) from [] (warn_slowpath_fmt+0x2c/0x3c) [ 13.525463] [] (warn_slowpath_fmt) from [] (check_flags+0x7c/0x1d0) [ 13.533447] [] (check_flags) from [] (lock_is_held+0x34/0x60) [ 13.540884] [] (lock_is_held) from [] (__schedule+0x98/0x7cc) [ 13.548342] [] (__schedule) from [] (do_work_pending+0x28/0xbc) [ 13.555962] [] (do_work_pending) from [] (work_pending+0xc/0x20) [ 13.563663] ---[ end trace 33bbfa4173cfa520 ]--- [ 13.568238] possible reason: unannotated irqs-off. [ 13.573005] irq event stamp: 5227 [ 13.576293] hardirqs last enabled at (5227): [] ret_fast_syscall+0x24/0x48 [ 13.584096] hardirqs last disabled at (5226): [] ret_fast_syscall+0xc/0x48 [ 13.591823] softirqs last enabled at (5222): [] __do_softirq+0x1e8/0x270 [ 13.599447] softirqs last disabled at (5205): [] irq_exit+0x84/0xf4 [ 13.656634] random: nonblocking pool is initialized Signed-off-by: Anand Moon --- arch/arm/kernel/signal.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c index bd19834..52bc9d6 100644 --- a/arch/arm/kernel/signal.c +++ b/arch/arm/kernel/signal.c @@ -476,7 +476,7 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs) /* * OK, we're invoking a handler - */ + */ static void handle_signal(struct ksignal *ksig, struct pt_regs *regs) { sigset_t *oldset = sigmask_to_save(); @@ -574,12 +574,14 @@ asmlinkage int do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) { do { + if (unlikely(!user_mode(regs))) + return 0; + /* Enable interrupts; they are disabled again on return to + * caller. */ + local_irq_enable(); if (likely(thread_flags & _TIF_NEED_RESCHED)) { schedule(); } else { - if (unlikely(!user_mode(regs))) - return 0; - local_irq_enable(); if (thread_flags & _TIF_SIGPENDING) { int restart = do_signal(regs, syscall); if (unlikely(restart)) { @@ -588,6 +590,7 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) * Deal with it without leaving * the kernel space. */ + local_irq_disable(); return restart; } syscall = 0; -- 2.1.0