linux-kernel - [PATCH v8 05/12] arm64: syscall: Rework el0_svc

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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20251126071446.3234218-6-ruanjinjie@huawei.com>
Date: Wed, 26 Nov 2025 15:14:39 +0800
From: Jinjie Ruan <ruanjinjie@...wei.com>
To: <catalin.marinas@....com>, <will@...nel.org>, <oleg@...hat.com>,
	<tglx@...utronix.de>, <peterz@...radead.org>, <luto@...nel.org>,
	<shuah@...nel.org>, <kees@...nel.org>, <wad@...omium.org>,
	<charlie@...osinc.com>, <akpm@...ux-foundation.org>, <ldv@...ace.io>,
	<macro@...am.me.uk>, <deller@....de>, <mark.rutland@....com>,
	<efault@....de>, <song@...nel.org>, <mbenes@...e.cz>, <ryan.roberts@....com>,
	<ada.coupriediaz@....com>, <anshuman.khandual@....com>, <broonie@...nel.org>,
	<kevin.brodsky@....com>, <pengcan@...inos.cn>, <dvyukov@...gle.com>,
	<kmal@...k.li>, <linux-arm-kernel@...ts.infradead.org>,
	<linux-kernel@...r.kernel.org>, <linux-kselftest@...r.kernel.org>
CC: <ruanjinjie@...wei.com>
Subject: [PATCH v8 05/12] arm64: syscall: Rework el0_svc_common()

The generic syscall syscall_exit_work() has the following content:

| audit_syscall_exit(regs)
| trace_sys_exit(regs, ...)
| ptrace_report_syscall_exit(regs, step)

The generic syscall syscall_exit_to_user_mode_work() has
the following form:

| unsigned long work = READ_ONCE(current_thread_info()->syscall_work)
| rseq_syscall()
| if (unlikely(work & SYSCALL_WORK_EXIT))
|	syscall_exit_work(regs, work)

In preparation for moving arm64 over to the generic entry code,
rework el0_svc_common() as below:

- Rename syscall_trace_exit() to syscall_exit_work().

- Add syscall_exit_to_user_mode_prepare() function to replace
  the combination of read_thread_flags() and syscall_exit_work(),
  also move the syscall exit check logic into it. Move has_syscall_work()
  helper into asm/syscall.h for reuse.

- As currently rseq_syscall() is always called and itself is controlled
  by the CONFIG_DEBUG_RSEQ macro, so the CONFIG_DEBUG_RSEQ check
  is removed.

Signed-off-by: Jinjie Ruan <ruanjinjie@...wei.com>
---
 arch/arm64/include/asm/syscall.h |  7 ++++++-
 arch/arm64/kernel/ptrace.c       | 14 +++++++++++---
 arch/arm64/kernel/syscall.c      | 20 +-------------------
 3 files changed, 18 insertions(+), 23 deletions(-)

diff --git a/arch/arm64/include/asm/syscall.h b/arch/arm64/include/asm/syscall.h
index d69f590a989b..6225981fbbdb 100644
--- a/arch/arm64/include/asm/syscall.h
+++ b/arch/arm64/include/asm/syscall.h
@@ -114,7 +114,12 @@ static inline int syscall_get_arch(struct task_struct *task)
 	return AUDIT_ARCH_AARCH64;
 }
 
+static inline bool has_syscall_work(unsigned long flags)
+{
+	return unlikely(flags & _TIF_SYSCALL_WORK);
+}
+
 int syscall_trace_enter(struct pt_regs *regs, long syscall, unsigned long flags);
-void syscall_trace_exit(struct pt_regs *regs, unsigned long flags);
+void syscall_exit_to_user_mode_prepare(struct pt_regs *regs);
 
 #endif	/* __ASM_SYSCALL_H */
diff --git a/arch/arm64/kernel/ptrace.c b/arch/arm64/kernel/ptrace.c
index dfdd886dc0a9..233a7688ac94 100644
--- a/arch/arm64/kernel/ptrace.c
+++ b/arch/arm64/kernel/ptrace.c
@@ -2409,10 +2409,8 @@ int syscall_trace_enter(struct pt_regs *regs, long syscall, unsigned long flags)
 	return syscall;
 }
 
-void syscall_trace_exit(struct pt_regs *regs, unsigned long flags)
+static void syscall_exit_work(struct pt_regs *regs, unsigned long flags)
 {
-	rseq_syscall(regs);
-
 	audit_syscall_exit(regs);
 
 	if (flags & _TIF_SYSCALL_TRACEPOINT)
@@ -2422,6 +2420,16 @@ void syscall_trace_exit(struct pt_regs *regs, unsigned long flags)
 		report_syscall_exit(regs);
 }
 
+void syscall_exit_to_user_mode_prepare(struct pt_regs *regs)
+{
+	unsigned long flags = read_thread_flags();
+
+	rseq_syscall(regs);
+
+	if (has_syscall_work(flags) || flags & _TIF_SINGLESTEP)
+		syscall_exit_work(regs, flags);
+}
+
 /*
  * SPSR_ELx bits which are always architecturally RES0 per ARM DDI 0487D.a.
  * We permit userspace to set SSBS (AArch64 bit 12, AArch32 bit 23) which is
diff --git a/arch/arm64/kernel/syscall.c b/arch/arm64/kernel/syscall.c
index ec31f82d2e9f..65021d0f49e1 100644
--- a/arch/arm64/kernel/syscall.c
+++ b/arch/arm64/kernel/syscall.c
@@ -65,11 +65,6 @@ static void invoke_syscall(struct pt_regs *regs, unsigned int scno,
 	choose_random_kstack_offset(get_random_u16());
 }
 
-static inline bool has_syscall_work(unsigned long flags)
-{
-	return unlikely(flags & _TIF_SYSCALL_WORK);
-}
-
 static void el0_svc_common(struct pt_regs *regs, int scno, int sc_nr,
 			   const syscall_fn_t syscall_table[])
 {
@@ -130,21 +125,8 @@ static void el0_svc_common(struct pt_regs *regs, int scno, int sc_nr,
 	}
 
 	invoke_syscall(regs, scno, sc_nr, syscall_table);
-
-	/*
-	 * The tracing status may have changed under our feet, so we have to
-	 * check again. However, if we were tracing entry, then we always trace
-	 * exit regardless, as the old entry assembly did.
-	 */
-	if (!has_syscall_work(flags) && !IS_ENABLED(CONFIG_DEBUG_RSEQ)) {
-		flags = read_thread_flags();
-		if (!has_syscall_work(flags) && !(flags & _TIF_SINGLESTEP))
-			return;
-	}
-
 trace_exit:
-	flags = read_thread_flags();
-	syscall_trace_exit(regs, flags);
+	syscall_exit_to_user_mode_prepare(regs);
 }
 
 void do_el0_svc(struct pt_regs *regs)
-- 
2.34.1