kernel.git/arch/s390/kernel/entry.S, branch linux-5.9.y

s390: fix irq state tracing

2020-12-16T09:58:34Z

[ Upstream commit b1cae1f84a0f609a34ebcaa087fbecef32f69882 ] With commit 58c644ba512c ("sched/idle: Fix arch_cpu_idle() vs tracing") common code calls arch_cpu_idle() with a lockdep state that tells irqs are on. This doesn't work very well for s390: psw_idle() will enable interrupts to wait for an interrupt. As soon as an interrupt occurs the interrupt handler will verify if the old context was psw_idle(). If that is the case the interrupt enablement bits in the old program status word will be cleared. A subsequent test in both the external as well as the io interrupt handler checks if in the old context interrupts were enabled. Due to the above patching of the old program status word it is assumed the old context had interrupts disabled, and therefore a call to TRACE_IRQS_OFF (aka trace_hardirqs_off_caller) is skipped. Which in turn makes lockdep incorrectly "think" that interrupts are enabled within the interrupt handler. Fix this by unconditionally calling TRACE_IRQS_OFF when entering interrupt handlers. Also call unconditionally TRACE_IRQS_ON when leaving interrupts handlers. This leaves the special psw_idle() case, which now returns with interrupts disabled, but has an "irqs on" lockdep state. So callers of psw_idle() must adjust the state on their own, if required. This is currently only __udelay_disabled(). Fixes: 58c644ba512c ("sched/idle: Fix arch_cpu_idle() vs tracing") Acked-by: Peter Zijlstra (Intel) Signed-off-by: Heiko Carstens Signed-off-by: Sasha Levin

s390: fix fpu restore in entry.S

2020-12-02T07:51:44Z

commit 1179f170b6f0af7bb0b3b7628136eaac450ddf31 upstream. We need to disable interrupts in load_fpu_regs(). Otherwise an interrupt might come in after the registers are loaded, but before CIF_FPU is cleared in load_fpu_regs(). When the interrupt returns, CIF_FPU will be cleared and the registers will never be restored. The entry.S code usually saves the interrupt state in __SF_EMPTY on the stack when disabling/restoring interrupts. sie64a however saves the pointer to the sie control block in __SF_SIE_CONTROL, which references the same location. This is non-obvious to the reader. To avoid thrashing the sie control block pointer in load_fpu_regs(), move the __SIE_* offsets eight bytes after __SF_EMPTY on the stack. Cc: # 5.8 Fixes: 0b0ed657fe00 ("s390: remove critical section cleanup from entry.S") Reported-by: Pierre Morel Signed-off-by: Sven Schnelle Acked-by: Christian Borntraeger Reviewed-by: Heiko Carstens Signed-off-by: Heiko Carstens Signed-off-by: Greg Kroah-Hartman

s390: fix system call exit path

2020-11-24T12:39:12Z

commit ce9dfafe29bed86fe3cda330ac6072ce84e1ff81 upstream. The system call exit path is running with interrupts enabled while checking for TIF/PIF/CIF bits which require special handling. If all bits have been checked interrupts are disabled and the kernel exits to user space. The problem is that after checking all bits and before interrupts are disabled bits can be set already again, due to interrupt handling. This means that the kernel can exit to user space with some TIF/PIF/CIF bits set, which should never happen. E.g. TIF_NEED_RESCHED might be set, which might lead to additional latencies, since that bit will only be recognized with next exit to user space. Fix this by checking the corresponding bits only when interrupts are disabled. Fixes: 0b0ed657fe00 ("s390: remove critical section cleanup from entry.S") Cc: # 5.8 Acked-by: Sven Schnelle Signed-off-by: Heiko Carstens Signed-off-by: Greg Kroah-Hartman

s390: fix comment regarding interrupts in svc

2020-07-10T13:08:23Z

With the removal of the critical section cleanup, we now enter the svc interrupt handler with interrupts disabled. Fixes: 0b0ed657fe00 ("s390: remove critical section cleanup from entry.S") Signed-off-by: Christian Borntraeger Signed-off-by: Heiko Carstens

s390: fix system call single stepping

2020-06-23T12:05:45Z

When single stepping an svc instruction on s390, the kernel is entered with a PER program check interruption. The program check handler than jumps to the system call handler by reloading the PSW. The code didn't set GPR13 to the thread pointer in struct task_struct. This made the kernel access invalid memory while trying to fetch the syscall function address. Fix this by always assigned GPR13 after .Lsysc_per. Fixes: 0b0ed657fe00 ("s390: remove critical section cleanup from entry.S") Reported-and-tested-by: Christian Borntraeger Signed-off-by: Sven Schnelle Reviewed-by: Heiko Carstens Signed-off-by: Heiko Carstens

s390/ptrace: pass invalid syscall numbers to tracing

2020-06-16T11:44:04Z

tracing expects to see invalid syscalls, so pass it through. The syscall path in entry.S checks the syscall number before looking up the handler, so it is still safe. Signed-off-by: Sven Schnelle Signed-off-by: Vasily Gorbik

s390: remove critical section cleanup from entry.S

2020-05-28T10:21:54Z

The current code is rather complex and caused a lot of subtle and hard to debug bugs in the past. Simplify the code by calling the system_call handler with interrupts disabled, save machine state, and re-enable them later. This requires significant changes to the machine check handling code as well. When the machine check interrupt arrived while being in kernel mode the new code will signal pending machine checks with a SIGP external call. When userspace was interrupted, the handler will switch to the kernel stack and directly execute s390_handle_mcck(). Signed-off-by: Sven Schnelle Signed-off-by: Vasily Gorbik

s390: prevent leaking kernel address in BEAR

2020-03-10T14:16:25Z

When userspace executes a syscall or gets interrupted, BEAR contains a kernel address when returning to userspace. This make it pretty easy to figure out where the kernel is mapped even with KASLR enabled. To fix this, add lpswe to lowcore and always execute it there, so userspace sees only the lowcore address of lpswe. For this we have to extend both critical_cleanup and the SWITCH_ASYNC macro to also check for lpswe addresses in lowcore. Fixes: b2d24b97b2a9 ("s390/kernel: add support for kernel address space layout randomization (KASLR)") Cc: # v5.2+ Reviewed-by: Gerald Schaefer Signed-off-by: Sven Schnelle Signed-off-by: Vasily Gorbik

sched/rt, s390: Use CONFIG_PREEMPTION

2019-12-08T13:37:35Z

CONFIG_PREEMPTION is selected by CONFIG_PREEMPT and by CONFIG_PREEMPT_RT. Both PREEMPT and PREEMPT_RT require the same functionality which today depends on CONFIG_PREEMPT. Switch the preemption and entry code over to use CONFIG_PREEMPTION. Add PREEMPT_RT output to die(). [bigeasy: +Kconfig, dumpstack.c] Signed-off-by: Thomas Gleixner Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Thomas Gleixner Cc: Christian Borntraeger Cc: Heiko Carstens Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Vasily Gorbik Cc: linux-s390@vger.kernel.org Link: https://lore.kernel.org/r/20191015191821.11479-18-bigeasy@linutronix.de Signed-off-by: Ingo Molnar

s390: enforce CONFIG_SMP

2019-06-07T08:09:37Z

There never have been distributions that shiped with CONFIG_SMP=n for s390. In addition the kernel currently doesn't even compile with CONFIG_SMP=n for s390. Most likely it wouldn't even work, even if we fix the compile error, since nobody tests it, since there is no use case that I can think of. Therefore simply enforce CONFIG_SMP and get rid of some more or less unused code. Reviewed-by: Christian Borntraeger Signed-off-by: Heiko Carstens