linux-kernel - Re: [PATCH v9 04/13] x86/um: nommu: syscall handling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <m24iw95kuf.wl-thehajime@gmail.com>
Date: Sat, 21 Jun 2025 21:40:24 +0900
From: Hajime Tazaki <thehajime@...il.com>
To: benjamin@...solutions.net
Cc: linux-um@...ts.infradead.org,
	ricarkol@...gle.com,
	Liam.Howlett@...cle.com,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v9 04/13] x86/um: nommu: syscall handling


On Thu, 19 Jun 2025 19:31:53 +0900,
Benjamin Berg wrote:

> > diff --git a/arch/x86/um/nommu/do_syscall_64.c b/arch/x86/um/nommu/do_syscall_64.c
> > new file mode 100644
> > index 000000000000..5d0fa83e7fdc
> > --- /dev/null
> > +++ b/arch/x86/um/nommu/do_syscall_64.c
> > @@ -0,0 +1,37 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/ptrace.h>
> > +#include <kern_util.h>
> > +#include <sysdep/syscalls.h>
> > +#include <os.h>
> > +
> > +__visible void do_syscall_64(struct pt_regs *regs)
> > +{
> > +	int syscall;
> > +
> > +	syscall = PT_SYSCALL_NR(regs->regs.gp);
> > +	UPT_SYSCALL_NR(&regs->regs) = syscall;
> > +
> > +	pr_debug("syscall(%d) (current=%lx) (fn=%lx)\n",
> > +		 syscall, (unsigned long)current,
> > +		 (unsigned long)sys_call_table[syscall]);
> 
> You probably want to drop the pr_debug from the syscall path.

okay, I'll update those parts.

> > +	if (likely(syscall < NR_syscalls)) {
> > +		PT_REGS_SET_SYSCALL_RETURN(regs,
> > +				EXECUTE_SYSCALL(syscall, regs));
> > +	}
> > +
> > +	pr_debug("syscall(%d) --> %lx\n", syscall,
> > +		regs->regs.gp[HOST_AX]);
> > +
> > +	PT_REGS_SYSCALL_RET(regs) = regs->regs.gp[HOST_AX];
> > +
> > +	/* execve succeeded */
> > +	if (syscall == __NR_execve && regs->regs.gp[HOST_AX] == 0)
> > +		userspace(&current->thread.regs.regs);
> 
> That said, this is what I am stumbling over. Why do you need to jump
> into userspace() here? It seems odd to me to need a special case in the
> syscall path itself.
> Aren't there other possibilities to hook/override the kernel task
> state?

thanks, I found that this is a leftover of our early implementation
which doesn't have a proper schedule upon an exit from syscall.  we
can remove this part and I'll fix them after more investigation.

> > +	/* force do_signal() --> is_syscall() */
> > +	set_thread_flag(TIF_SIGPENDING);
> > +	interrupt_end();
> 
> Same here. The MMU UML code seems to also do this, but restricted to
> ptrace'd processes? Maybe I am just missing something obvious …

nommu doesn't have separate process/context to indicate a schedule to
the context here (syscall).  without that part we do not have a chance
to schedule tasks and signals to userspace.

But the force on SIGPENDING flag is not actually needed so, I'll
remove that part.

thanks for pointing out.

-- Hajime