[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250111011632.GA1724@strace.io>
Date: Sat, 11 Jan 2025 03:16:32 +0200
From: "Dmitry V. Levin" <ldv@...ace.io>
To: Sven Schnelle <svens@...ux.ibm.com>
Cc: Oleg Nesterov <oleg@...hat.com>,
Eugene Syromyatnikov <evgsyr@...il.com>,
Mike Frysinger <vapier@...too.org>,
Renzo Davoli <renzo@...unibo.it>,
Davide Berardi <berardi.dav@...il.com>,
strace-devel@...ts.strace.io, Vineet Gupta <vgupta@...nel.org>,
Russell King <linux@...linux.org.uk>,
Catalin Marinas <catalin.marinas@....com>,
Will Deacon <will@...nel.org>, Brian Cain <bcain@...cinc.com>,
Huacai Chen <chenhuacai@...nel.org>,
WANG Xuerui <kernel@...0n.name>,
Geert Uytterhoeven <geert@...ux-m68k.org>,
Michal Simek <monstr@...str.eu>,
Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
Dinh Nguyen <dinguyen@...nel.org>, Jonas Bonn <jonas@...thpole.se>,
Stefan Kristiansson <stefan.kristiansson@...nalahti.fi>,
Stafford Horne <shorne@...il.com>,
"James E.J. Bottomley" <James.Bottomley@...senPartnership.com>,
Helge Deller <deller@....de>, Michael Ellerman <mpe@...erman.id.au>,
Nicholas Piggin <npiggin@...il.com>,
Christophe Leroy <christophe.leroy@...roup.eu>,
Naveen N Rao <naveen@...nel.org>,
Madhavan Srinivasan <maddy@...ux.ibm.com>,
Paul Walmsley <paul.walmsley@...ive.com>,
Palmer Dabbelt <palmer@...belt.com>,
Albert Ou <aou@...s.berkeley.edu>,
Heiko Carstens <hca@...ux.ibm.com>,
Vasily Gorbik <gor@...ux.ibm.com>,
Alexander Gordeev <agordeev@...ux.ibm.com>,
Christian Borntraeger <borntraeger@...ux.ibm.com>,
Yoshinori Sato <ysato@...rs.sourceforge.jp>,
Rich Felker <dalias@...c.org>,
John Paul Adrian Glaubitz <glaubitz@...sik.fu-berlin.de>,
"David S. Miller" <davem@...emloft.net>,
Andreas Larsson <andreas@...sler.com>,
Richard Weinberger <richard@....at>,
Anton Ivanov <anton.ivanov@...bridgegreys.com>,
Johannes Berg <johannes@...solutions.net>,
Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
"H. Peter Anvin" <hpa@...or.com>, Chris Zankel <chris@...kel.net>,
Max Filippov <jcmvbkbc@...il.com>, Arnd Bergmann <arnd@...db.de>,
linux-snps-arc@...ts.infradead.org, linux-kernel@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org, linux-hexagon@...r.kernel.org,
loongarch@...ts.linux.dev, linux-m68k@...ts.linux-m68k.org,
linux-mips@...r.kernel.org, linux-openrisc@...r.kernel.org,
linux-parisc@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
linux-riscv@...ts.infradead.org, linux-s390@...r.kernel.org,
linux-sh@...r.kernel.org, sparclinux@...r.kernel.org,
linux-um@...ts.infradead.org, linux-arch@...r.kernel.org
Subject: Re: [PATCH 3/6] syscall.h: introduce syscall_set_nr()
On Fri, Jan 10, 2025 at 08:37:46AM +0100, Sven Schnelle wrote:
> "Dmitry V. Levin" <ldv@...ace.io> writes:
>
> > Similar to syscall_set_arguments() that complements
> > syscall_get_arguments(), introduce syscall_set_nr()
> > that complements syscall_get_nr().
> >
> > syscall_set_nr() is going to be needed along with
> > syscall_set_arguments() on all HAVE_ARCH_TRACEHOOK
> > architectures to implement PTRACE_SET_SYSCALL_INFO API.
[...]
> > diff --git a/arch/s390/include/asm/syscall.h b/arch/s390/include/asm/syscall.h
> > index b3dd883699e7..1c0e349fd5c9 100644
> > --- a/arch/s390/include/asm/syscall.h
> > +++ b/arch/s390/include/asm/syscall.h
> > @@ -24,6 +24,13 @@ static inline long syscall_get_nr(struct task_struct *task,
> > (regs->int_code & 0xffff) : -1;
> > }
> >
> > +static inline void syscall_set_nr(struct task_struct *task,
> > + struct pt_regs *regs,
> > + int nr)
> > +{
>
> I think there should be a
>
> if (!test_pt_regs_flags(regs, PIF_SYSCALL))
> return;
>
> before the modification so a user can't accidentally change int_code
> when ptrace stopped in a non-syscall path.
The reason why syscall_get_nr() has this check on s390 (and similar checks
on arc, powerpc, and sparc) is that syscall_get_nr() can be called while
the target task is not in syscall.
Unlike syscall_get_nr(), syscall_set_nr() can be called only when the
target task is stopped for tracing on entering syscall: the description in
include/asm-generic/syscall.h explicitly states that, and the follow-up
patch that introduces PTRACE_SET_SYSCALL_INFO adds a syscall_set_nr() call
when the tracee is stopped on entering syscall in either
PTRACE_SYSCALL_INFO_ENTRY or PTRACE_SYSCALL_INFO_SECCOMP state.
I don't mind adding a check, but syscall_set_nr() invocation while the
target task is not in syscall wouldn't be a result of user actions but
a kernel programing error, and in that case WARN_ON_ONCE() would be more
appropriate.
If calling syscall_set_nr() while the target task is not in syscall was
legal, then syscall_set_nr() would have been designed to return a value
indicating the status of operation.
Anyway, I'll add an explanatory comment to syscall_set_nr() on all
architectures where syscall_get_nr() has a check.
--
ldv
Powered by blists - more mailing lists