[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH v9 16/36] x86/fred: Allow single-step trap and NMI when starting a new task
- To: linux-doc@xxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, linux-edac@xxxxxxxxxxxxxxx, linux-hyperv@xxxxxxxxxxxxxxx, kvm@xxxxxxxxxxxxxxx, xen-devel@xxxxxxxxxxxxxxxxxxxx
- From: Xin Li <xin3.li@xxxxxxxxx>
- Date: Sun, 30 Jul 2023 23:32:57 -0700
- Cc: Jonathan Corbet <corbet@xxxxxxx>, Thomas Gleixner <tglx@xxxxxxxxxxxxx>, Ingo Molnar <mingo@xxxxxxxxxx>, Borislav Petkov <bp@xxxxxxxxx>, Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>, x86@xxxxxxxxxx, "H . Peter Anvin" <hpa@xxxxxxxxx>, Andy Lutomirski <luto@xxxxxxxxxx>, Oleg Nesterov <oleg@xxxxxxxxxx>, Tony Luck <tony.luck@xxxxxxxxx>, "K . Y . Srinivasan" <kys@xxxxxxxxxxxxx>, Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>, Wei Liu <wei.liu@xxxxxxxxxx>, Dexuan Cui <decui@xxxxxxxxxxxxx>, Paolo Bonzini <pbonzini@xxxxxxxxxx>, Wanpeng Li <wanpengli@xxxxxxxxxxx>, Vitaly Kuznetsov <vkuznets@xxxxxxxxxx>, Sean Christopherson <seanjc@xxxxxxxxxx>, Peter Zijlstra <peterz@xxxxxxxxxxxxx>, Juergen Gross <jgross@xxxxxxxx>, Stefano Stabellini <sstabellini@xxxxxxxxxx>, Oleksandr Tyshchenko <oleksandr_tyshchenko@xxxxxxxx>, Josh Poimboeuf <jpoimboe@xxxxxxxxxx>, "Paul E . McKenney" <paulmck@xxxxxxxxxx>, Catalin Marinas <catalin.marinas@xxxxxxx>, Randy Dunlap <rdunlap@xxxxxxxxxxxxx>, Steven Rostedt <rostedt@xxxxxxxxxxx>, Kim Phillips <kim.phillips@xxxxxxx>, Xin Li <xin3.li@xxxxxxxxx>, Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx>, "Liam R . Howlett" <Liam.Howlett@xxxxxxxxxx>, Sebastian Reichel <sebastian.reichel@xxxxxxxxxxxxx>, "Kirill A . Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>, Suren Baghdasaryan <surenb@xxxxxxxxxx>, Pawan Gupta <pawan.kumar.gupta@xxxxxxxxxxxxxxx>, Jiaxi Chen <jiaxi.chen@xxxxxxxxxxxxxxx>, Babu Moger <babu.moger@xxxxxxx>, Jim Mattson <jmattson@xxxxxxxxxx>, Sandipan Das <sandipan.das@xxxxxxx>, Lai Jiangshan <jiangshanlai@xxxxxxxxx>, Hans de Goede <hdegoede@xxxxxxxxxx>, Reinette Chatre <reinette.chatre@xxxxxxxxx>, Daniel Sneddon <daniel.sneddon@xxxxxxxxxxxxxxx>, Breno Leitao <leitao@xxxxxxxxxx>, Nikunj A Dadhania <nikunj@xxxxxxx>, Brian Gerst <brgerst@xxxxxxxxx>, Sami Tolvanen <samitolvanen@xxxxxxxxxx>, Alexander Potapenko <glider@xxxxxxxxxx>, Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>, Arnd Bergmann <arnd@xxxxxxxx>, "Eric W . Biederman" <ebiederm@xxxxxxxxxxxx>, Kees Cook <keescook@xxxxxxxxxxxx>, Masami Hiramatsu <mhiramat@xxxxxxxxxx>, Masahiro Yamada <masahiroy@xxxxxxxxxx>, Ze Gao <zegao2021@xxxxxxxxx>, Fei Li <fei1.li@xxxxxxxxx>, Conghui <conghui.chen@xxxxxxxxx>, Ashok Raj <ashok.raj@xxxxxxxxx>, "Jason A . Donenfeld" <Jason@xxxxxxxxx>, Mark Rutland <mark.rutland@xxxxxxx>, Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>, Jiapeng Chong <jiapeng.chong@xxxxxxxxxxxxxxxxx>, Jane Malalane <jane.malalane@xxxxxxxxxx>, David Woodhouse <dwmw@xxxxxxxxxxxx>, Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>, Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>, Yantengsi <siyanteng@xxxxxxxxxxx>, Christophe Leroy <christophe.leroy@xxxxxxxxxx>, Sathvika Vasireddy <sv@xxxxxxxxxxxxx>
- Delivery-date: Mon, 31 Jul 2023 07:03:10 +0000
- List-id: Xen developer discussion <xen-devel.lists.xenproject.org>
From: "H. Peter Anvin (Intel)" <hpa@xxxxxxxxx>
Entering a new task is logically speaking a return from a system call
(exec, fork, clone, etc.). As such, if ptrace enables single stepping
a single step exception should be allowed to trigger immediately upon
entering user space. This is not optional.
NMI should *never* be disabled in user space. As such, this is an
optional, opportunistic way to catch errors.
Allow single-step trap and NMI when starting a new task, thus once
the new task enters user space, single-step trap and NMI are both
enabled immediately.
Signed-off-by: H. Peter Anvin (Intel) <hpa@xxxxxxxxx>
Tested-by: Shan Kang <shan.kang@xxxxxxxxx>
Signed-off-by: Xin Li <xin3.li@xxxxxxxxx>
---
Changes since v8:
* Use high-order 48 bits above the lowest 16 bit SS only when FRED
is enabled (Thomas Gleixner).
---
arch/x86/kernel/process_64.c | 23 +++++++++++++++++------
1 file changed, 17 insertions(+), 6 deletions(-)
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 6d5fed29f552..0b47871a6141 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -56,6 +56,7 @@
#include <asm/resctrl.h>
#include <asm/unistd.h>
#include <asm/fsgsbase.h>
+#include <asm/fred.h>
#ifdef CONFIG_IA32_EMULATION
/* Not included via unistd.h */
#include <asm/unistd_32_ia32.h>
@@ -507,8 +508,18 @@ void x86_gsbase_write_task(struct task_struct *task,
unsigned long gsbase)
static void
start_thread_common(struct pt_regs *regs, unsigned long new_ip,
unsigned long new_sp,
- unsigned int _cs, unsigned int _ss, unsigned int _ds)
+ u16 _cs, u16 _ss, u16 _ds)
{
+ /*
+ * Paranoia: High-order 48 bits above the lowest 16 bit SS are
+ * discarded by the legacy IRET instruction on all Intel, AMD,
+ * and Cyrix/Centaur/VIA CPUs, thus can be set unconditionally,
+ * even when FRED is not enabled. But we choose the safer side
+ * to use these bits only when FRED is enabled.
+ */
+ const unsigned long ssx_flags = cpu_feature_enabled(X86_FEATURE_FRED) ?
+ (FRED_SSX_SOFTWARE_INITIATED | FRED_SSX_NMI) : 0;
+
WARN_ON_ONCE(regs != current_pt_regs());
if (static_cpu_has(X86_BUG_NULL_SEG)) {
@@ -522,11 +533,11 @@ start_thread_common(struct pt_regs *regs, unsigned long
new_ip,
loadsegment(ds, _ds);
load_gs_index(0);
- regs->ip = new_ip;
- regs->sp = new_sp;
- regs->cs = _cs;
- regs->ss = _ss;
- regs->flags = X86_EFLAGS_IF;
+ regs->ip = new_ip;
+ regs->sp = new_sp;
+ regs->csx = _cs;
+ regs->ssx = _ss | ssx_flags;
+ regs->flags = X86_EFLAGS_IF | X86_EFLAGS_FIXED;
}
void
--
2.34.1
|