linux-yocto/arch/um/kernel/skas/syscall.c
Benjamin Berg 887c5c12e8 um: work around sched_yield not yielding in time-travel mode
sched_yield by a userspace may not actually cause scheduling in
time-travel mode as no time has passed. In the case seen it appears to
be a badly implemented userspace spinlock in ASAN. Unfortunately, with
time-travel it causes an extreme slowdown or even deadlock depending on
the kernel configuration (CONFIG_UML_MAX_USERSPACE_ITERATIONS).

Work around it by accounting time to the process whenever it executes a
sched_yield syscall.

Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Link: https://patch.msgid.link/20250314130815.226872-1-benjamin@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2025-03-18 11:06:20 +01:00

68 lines
1.7 KiB
C

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
*/
#include <linux/kernel.h>
#include <linux/ptrace.h>
#include <linux/seccomp.h>
#include <kern_util.h>
#include <sysdep/ptrace.h>
#include <sysdep/ptrace_user.h>
#include <sysdep/syscalls.h>
#include <linux/time-internal.h>
#include <asm/unistd.h>
#include <asm/delay.h>
void handle_syscall(struct uml_pt_regs *r)
{
struct pt_regs *regs = container_of(r, struct pt_regs, regs);
int syscall;
/* Initialize the syscall number and default return value. */
UPT_SYSCALL_NR(r) = PT_SYSCALL_NR(r->gp);
PT_REGS_SET_SYSCALL_RETURN(regs, -ENOSYS);
if (syscall_trace_enter(regs))
goto out;
/* Do the seccomp check after ptrace; failures should be fast. */
if (secure_computing() == -1)
goto out;
syscall = UPT_SYSCALL_NR(r);
/*
* If no time passes, then sched_yield may not actually yield, causing
* broken spinlock implementations in userspace (ASAN) to hang for long
* periods of time.
*/
if ((time_travel_mode == TT_MODE_INFCPU ||
time_travel_mode == TT_MODE_EXTERNAL) &&
syscall == __NR_sched_yield)
tt_extra_sched_jiffies += 1;
if (syscall >= 0 && syscall < __NR_syscalls) {
unsigned long ret = EXECUTE_SYSCALL(syscall, regs);
PT_REGS_SET_SYSCALL_RETURN(regs, ret);
/*
* An error value here can be some form of -ERESTARTSYS
* and then we'd just loop. Make any error syscalls take
* some time, so that it won't just loop if something is
* not ready, and hopefully other things will make some
* progress.
*/
if (IS_ERR_VALUE(ret) &&
(time_travel_mode == TT_MODE_INFCPU ||
time_travel_mode == TT_MODE_EXTERNAL)) {
um_udelay(1);
schedule();
}
}
out:
syscall_trace_leave(regs);
}