Merge tag 'x86-entry-2020-03-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 entry code updates from Thomas Gleixner:

 - Convert the 32bit syscalls to be pt_regs based which removes the
   requirement to push all 6 potential arguments onto the stack and
   consolidates the interface with the 64bit variant

 - The first small portion of the exception and syscall related entry
   code consolidation which aims to address the recently discovered
   issues vs. RCU, int3, NMI and some other exceptions which can
   interrupt any context. The bulk of the changes is still work in
   progress and aimed for 5.8.

 - A few lockdep namespace cleanups which have been applied into this
   branch to keep the prerequisites for the ongoing work confined.

* tag 'x86-entry-2020-03-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (35 commits)
  x86/entry: Fix build error x86 with !CONFIG_POSIX_TIMERS
  lockdep: Rename trace_{hard,soft}{irq_context,irqs_enabled}()
  lockdep: Rename trace_softirqs_{on,off}()
  lockdep: Rename trace_hardirq_{enter,exit}()
  x86/entry: Rename ___preempt_schedule
  x86: Remove unneeded includes
  x86/entry: Drop asmlinkage from syscalls
  x86/entry/32: Enable pt_regs based syscalls
  x86/entry/32: Use IA32-specific wrappers for syscalls taking 64-bit arguments
  x86/entry/32: Rename 32-bit specific syscalls
  x86/entry/32: Clean up syscall_32.tbl
  x86/entry: Remove ABI prefixes from functions in syscall tables
  x86/entry/64: Add __SYSCALL_COMMON()
  x86/entry: Remove syscall qualifier support
  x86/entry/64: Remove ptregs qualifier from syscall table
  x86/entry: Move max syscall number calculation to syscallhdr.sh
  x86/entry/64: Split X32 syscall table into its own file
  x86/entry/64: Move sys_ni_syscall stub to common.c
  x86/entry/64: Use syscall wrappers for x32_rt_sigreturn
  x86/entry: Refactor SYS_NI macros
  ...

arch/x86/Kconfig

@@ -30,7 +30,6 @@ config X86_64
 	select MODULES_USE_ELF_RELA
 	select NEED_DMA_MAP_STATE
 	select SWIOTLB
-	select ARCH_HAS_SYSCALL_WRAPPER
 
 config FORCE_DYNAMIC_FTRACE
 	def_bool y
@@ -79,6 +78,7 @@ config X86
 	select ARCH_HAS_STRICT_KERNEL_RWX
 	select ARCH_HAS_STRICT_MODULE_RWX
 	select ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
+	select ARCH_HAS_SYSCALL_WRAPPER
 	select ARCH_HAS_UBSAN_SANITIZE_ALL
 	select ARCH_HAVE_NMI_SAFE_CMPXCHG
 	select ARCH_MIGHT_HAVE_ACPI_PDC		if ACPI

arch/x86/entry/Makefile

@@ -14,4 +14,5 @@ obj-y				+= vdso/
 obj-y				+= vsyscall/
 
 obj-$(CONFIG_IA32_EMULATION)	+= entry_64_compat.o syscall_32.o
+obj-$(CONFIG_X86_X32_ABI)	+= syscall_x32.o
 

arch/x86/entry/common.c

@@ -333,20 +333,7 @@ static __always_inline void do_syscall_32_irqs_on(struct pt_regs *regs)
 
 	if (likely(nr < IA32_NR_syscalls)) {
 		nr = array_index_nospec(nr, IA32_NR_syscalls);
-#ifdef CONFIG_IA32_EMULATION
 		regs->ax = ia32_sys_call_table[nr](regs);
-#else
-		/*
-		 * It's possible that a 32-bit syscall implementation
-		 * takes a 64-bit parameter but nonetheless assumes that
-		 * the high bits are zero.  Make sure we zero-extend all
-		 * of the args.
-		 */
-		regs->ax = ia32_sys_call_table[nr](
-			(unsigned int)regs->bx, (unsigned int)regs->cx,
-			(unsigned int)regs->dx, (unsigned int)regs->si,
-			(unsigned int)regs->di, (unsigned int)regs->bp);
-#endif /* CONFIG_IA32_EMULATION */
 	}
 
 	syscall_return_slowpath(regs);
@@ -438,3 +425,8 @@ __visible long do_fast_syscall_32(struct pt_regs *regs)
 #endif
 }
 #endif
+
+SYSCALL_DEFINE0(ni_syscall)
+{
+	return -ENOSYS;
+}

arch/x86/entry/entry_32.S

@@ -1088,10 +1088,10 @@ SYM_FUNC_START(entry_INT80_32)
 	STACKLEAK_ERASE
 
 restore_all:
-	TRACE_IRQS_IRET
+	TRACE_IRQS_ON
 	SWITCH_TO_ENTRY_STACK
 	CHECK_AND_APPLY_ESPFIX
-.Lrestore_nocheck:
+
 	/* Switch back to user CR3 */
 	SWITCH_TO_USER_CR3 scratch_reg=%eax
 
@@ -1290,7 +1290,7 @@ SYM_CODE_END(simd_coprocessor_error)
 
 SYM_CODE_START(device_not_available)
 	ASM_CLAC
-	pushl	$-1				# mark this as an int
+	pushl	$0
 	pushl	$do_device_not_available
 	jmp	common_exception
 SYM_CODE_END(device_not_available)
@@ -1365,7 +1365,7 @@ SYM_CODE_END(divide_error)
 SYM_CODE_START(machine_check)
 	ASM_CLAC
 	pushl	$0
-	pushl	machine_check_vector
+	pushl	$do_mce
 	jmp	common_exception
 SYM_CODE_END(machine_check)
 #endif
@@ -1531,7 +1531,7 @@ SYM_CODE_START(debug)
 	 * Entry from sysenter is now handled in common_exception
 	 */
 	ASM_CLAC
-	pushl	$-1				# mark this as an int
+	pushl	$0
 	pushl	$do_debug
 	jmp	common_exception
 SYM_CODE_END(debug)
@@ -1682,18 +1682,13 @@ SYM_CODE_END(nmi)
 
 SYM_CODE_START(int3)
 	ASM_CLAC
-	pushl	$-1				# mark this as an int
-
-	SAVE_ALL switch_stacks=1
-	ENCODE_FRAME_POINTER
-	TRACE_IRQS_OFF
-	xorl	%edx, %edx			# zero error code
-	movl	%esp, %eax			# pt_regs pointer
-	call	do_int3
-	jmp	ret_from_exception
+	pushl	$0
+	pushl	$do_int3
+	jmp	common_exception
 SYM_CODE_END(int3)
 
 SYM_CODE_START(general_protection)
+	ASM_CLAC
 	pushl	$do_general_protection
 	jmp	common_exception
 SYM_CODE_END(general_protection)

arch/x86/entry/entry_64.S

@@ -174,7 +174,7 @@ SYM_INNER_LABEL(entry_SYSCALL_64_after_hwframe, SYM_L_GLOBAL)
 	movq	%rsp, %rsi
 	call	do_syscall_64		/* returns with IRQs disabled */
 
-	TRACE_IRQS_IRETQ		/* we're about to change IF */
+	TRACE_IRQS_ON			/* return enables interrupts */
 
 	/*
 	 * Try to use SYSRET instead of IRET if we're returning to
@@ -619,7 +619,7 @@ ret_from_intr:
 .Lretint_user:
 	mov	%rsp,%rdi
 	call	prepare_exit_to_usermode
-	TRACE_IRQS_IRETQ
+	TRACE_IRQS_ON
 
 SYM_INNER_LABEL(swapgs_restore_regs_and_return_to_usermode, SYM_L_GLOBAL)
 #ifdef CONFIG_DEBUG_ENTRY

arch/x86/entry/syscall_32.c

@@ -4,29 +4,22 @@
 #include <linux/linkage.h>
 #include <linux/sys.h>
 #include <linux/cache.h>
-#include <asm/asm-offsets.h>
+#include <linux/syscalls.h>
+#include <asm/unistd.h>
 #include <asm/syscall.h>
 
-#ifdef CONFIG_IA32_EMULATION
-/* On X86_64, we use struct pt_regs * to pass parameters to syscalls */
-#define __SYSCALL_I386(nr, sym, qual) extern asmlinkage long sym(const struct pt_regs *);
-#define __sys_ni_syscall __ia32_sys_ni_syscall
-#else /* CONFIG_IA32_EMULATION */
-#define __SYSCALL_I386(nr, sym, qual) extern asmlinkage long sym(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
-extern asmlinkage long sys_ni_syscall(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
-#define __sys_ni_syscall sys_ni_syscall
-#endif /* CONFIG_IA32_EMULATION */
+#define __SYSCALL_I386(nr, sym) extern long __ia32_##sym(const struct pt_regs *);
 
 #include <asm/syscalls_32.h>
 #undef __SYSCALL_I386
 
-#define __SYSCALL_I386(nr, sym, qual) [nr] = sym,
+#define __SYSCALL_I386(nr, sym) [nr] = __ia32_##sym,
 
-__visible const sys_call_ptr_t ia32_sys_call_table[__NR_syscall_compat_max+1] = {
+__visible const sys_call_ptr_t ia32_sys_call_table[__NR_ia32_syscall_max+1] = {
 	/*
 	 * Smells like a compiler bug -- it doesn't work
 	 * when the & below is removed.
 	 */
-	[0 ... __NR_syscall_compat_max] = &__sys_ni_syscall,
+	[0 ... __NR_ia32_syscall_max] = &__ia32_sys_ni_syscall,
 #include <asm/syscalls_32.h>
 };

arch/x86/entry/syscall_64.c

@@ -5,24 +5,17 @@
 #include <linux/sys.h>
 #include <linux/cache.h>
 #include <linux/syscalls.h>
-#include <asm/asm-offsets.h>
+#include <asm/unistd.h>
 #include <asm/syscall.h>
 
-extern asmlinkage long sys_ni_syscall(void);
+#define __SYSCALL_X32(nr, sym)
+#define __SYSCALL_COMMON(nr, sym) __SYSCALL_64(nr, sym)
 
-SYSCALL_DEFINE0(ni_syscall)
-{
-	return sys_ni_syscall();
-}
-
-#define __SYSCALL_64(nr, sym, qual) extern asmlinkage long sym(const struct pt_regs *);
-#define __SYSCALL_X32(nr, sym, qual) __SYSCALL_64(nr, sym, qual)
+#define __SYSCALL_64(nr, sym) extern long __x64_##sym(const struct pt_regs *);
 #include <asm/syscalls_64.h>
 #undef __SYSCALL_64
-#undef __SYSCALL_X32
 
-#define __SYSCALL_64(nr, sym, qual) [nr] = sym,
-#define __SYSCALL_X32(nr, sym, qual)
+#define __SYSCALL_64(nr, sym) [nr] = __x64_##sym,
 
 asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
 	/*
@@ -32,25 +25,3 @@ asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
 	[0 ... __NR_syscall_max] = &__x64_sys_ni_syscall,
 #include <asm/syscalls_64.h>
 };
-
-#undef __SYSCALL_64
-#undef __SYSCALL_X32
-
-#ifdef CONFIG_X86_X32_ABI
-
-#define __SYSCALL_64(nr, sym, qual)
-#define __SYSCALL_X32(nr, sym, qual) [nr] = sym,
-
-asmlinkage const sys_call_ptr_t x32_sys_call_table[__NR_syscall_x32_max+1] = {
-	/*
-	 * Smells like a compiler bug -- it doesn't work
-	 * when the & below is removed.
-	 */
-	[0 ... __NR_syscall_x32_max] = &__x64_sys_ni_syscall,
-#include <asm/syscalls_64.h>
-};
-
-#undef __SYSCALL_64
-#undef __SYSCALL_X32
-
-#endif

arch/x86/entry/syscall_x32.c

+// SPDX-License-Identifier: GPL-2.0
+/* System call table for x32 ABI. */
+
+#include <linux/linkage.h>
+#include <linux/sys.h>
+#include <linux/cache.h>
+#include <linux/syscalls.h>
+#include <asm/unistd.h>
+#include <asm/syscall.h>
+
+#define __SYSCALL_64(nr, sym)
+
+#define __SYSCALL_X32(nr, sym) extern long __x32_##sym(const struct pt_regs *);
+#define __SYSCALL_COMMON(nr, sym) extern long __x64_##sym(const struct pt_regs *);
+#include <asm/syscalls_64.h>
+#undef __SYSCALL_X32
+#undef __SYSCALL_COMMON
+
+#define __SYSCALL_X32(nr, sym) [nr] = __x32_##sym,
+#define __SYSCALL_COMMON(nr, sym) [nr] = __x64_##sym,
+
+asmlinkage const sys_call_ptr_t x32_sys_call_table[__NR_x32_syscall_max+1] = {
+	/*
+	 * Smells like a compiler bug -- it doesn't work
+	 * when the & below is removed.
+	 */
+	[0 ... __NR_x32_syscall_max] = &__x64_sys_ni_syscall,
+#include <asm/syscalls_64.h>
+};

arch/x86/entry/syscalls/syscallhdr.sh

@@ -15,14 +15,21 @@ grep -E "^[0-9A-Fa-fXx]+[[:space:]]+${my_abis}" "$in" | sort -n | (
     echo "#define ${fileguard} 1"
     echo ""
 
+    max=0
     while read nr abi name entry ; do
 	if [ -z "$offset" ]; then
 	    echo "#define __NR_${prefix}${name} $nr"
 	else
 	    echo "#define __NR_${prefix}${name} ($offset + $nr)"
         fi
+
+	max=$nr
     done
 
+    echo ""
+    echo "#ifdef __KERNEL__"
+    echo "#define __NR_${prefix}syscall_max $max"
+    echo "#endif"
     echo ""
     echo "#endif /* ${fileguard} */"
 ) > "$out"

arch/x86/entry/syscalls/syscalltbl.sh

@@ -9,15 +9,7 @@ syscall_macro() {
     local nr="$2"
     local entry="$3"
 
-    # Entry can be either just a function name or "function/qualifier"
-    real_entry="${entry%%/*}"
-    if [ "$entry" = "$real_entry" ]; then
-        qualifier=
-    else
-        qualifier=${entry#*/}
-    fi
-
-    echo "__SYSCALL_${abi}($nr, $real_entry, $qualifier)"
+    echo "__SYSCALL_${abi}($nr, $entry)"
 }
 
 emit() {
@@ -25,27 +17,15 @@ emit() {
     local nr="$2"
     local entry="$3"
     local compat="$4"
-    local umlentry=""
 
     if [ "$abi" != "I386" -a -n "$compat" ]; then
 	echo "a compat entry ($abi: $compat) for a 64-bit syscall makes no sense" >&2
 	exit 1
     fi
 
-    # For CONFIG_UML, we need to strip the __x64_sys prefix
-    if [ "$abi" = "64" -a "${entry}" != "${entry#__x64_sys}" ]; then
-	    umlentry="sys${entry#__x64_sys}"
-    fi
-
     if [ -z "$compat" ]; then
-	if [ -n "$entry" -a -z "$umlentry" ]; then
-	    syscall_macro "$abi" "$nr" "$entry"
-	elif [ -n "$umlentry" ]; then # implies -n "$entry"
-	    echo "#ifdef CONFIG_X86"
+	if [ -n "$entry" ]; then
 	    syscall_macro "$abi" "$nr" "$entry"
-	    echo "#else /* CONFIG_UML */"
-	    syscall_macro "$abi" "$nr" "$umlentry"
-	    echo "#endif"
 	fi
     else
 	echo "#ifdef CONFIG_X86_32"
@@ -61,24 +41,6 @@ emit() {
 grep '^[0-9]' "$in" | sort -n | (
     while read nr abi name entry compat; do
 	abi=`echo "$abi" | tr '[a-z]' '[A-Z]'`
-	if [ "$abi" = "COMMON" -o "$abi" = "64" ]; then
-	    emit 64 "$nr" "$entry" "$compat"
-	    if [ "$abi" = "COMMON" ]; then
-		# COMMON means that this syscall exists in the same form for
-		# 64-bit and X32.
-		echo "#ifdef CONFIG_X86_X32_ABI"
-		emit X32 "$nr" "$entry" "$compat"
-		echo "#endif"
-	    fi
-	elif [ "$abi" = "X32" ]; then
-	    echo "#ifdef CONFIG_X86_X32_ABI"
-	    emit X32 "$nr" "$entry" "$compat"
-	    echo "#endif"
-	elif [ "$abi" = "I386" ]; then
-	    emit "$abi" "$nr" "$entry" "$compat"
-	else
-	    echo "Unknown abi $abi" >&2
-	    exit 1
-	fi
+	emit "$abi" "$nr" "$entry" "$compat"
     done
 ) > "$out"

arch/x86/entry/thunk_32.S

@@ -35,9 +35,9 @@ SYM_CODE_END(\name)
 #endif
 
 #ifdef CONFIG_PREEMPTION
-	THUNK ___preempt_schedule, preempt_schedule
-	THUNK ___preempt_schedule_notrace, preempt_schedule_notrace
-	EXPORT_SYMBOL(___preempt_schedule)
-	EXPORT_SYMBOL(___preempt_schedule_notrace)
+	THUNK preempt_schedule_thunk, preempt_schedule
+	THUNK preempt_schedule_notrace_thunk, preempt_schedule_notrace
+	EXPORT_SYMBOL(preempt_schedule_thunk)
+	EXPORT_SYMBOL(preempt_schedule_notrace_thunk)
 #endif
 

arch/x86/entry/thunk_64.S

@@ -47,10 +47,10 @@ SYM_FUNC_END(\name)
 #endif
 
 #ifdef CONFIG_PREEMPTION
-	THUNK ___preempt_schedule, preempt_schedule
-	THUNK ___preempt_schedule_notrace, preempt_schedule_notrace
-	EXPORT_SYMBOL(___preempt_schedule)
-	EXPORT_SYMBOL(___preempt_schedule_notrace)
+	THUNK preempt_schedule_thunk, preempt_schedule
+	THUNK preempt_schedule_notrace_thunk, preempt_schedule_notrace
+	EXPORT_SYMBOL(preempt_schedule_thunk)
+	EXPORT_SYMBOL(preempt_schedule_notrace_thunk)
 #endif
 
 #if defined(CONFIG_TRACE_IRQFLAGS) \

arch/x86/entry/vdso/vdso32/vclock_gettime.c

@@ -13,6 +13,7 @@
  */
 #undef CONFIG_64BIT
 #undef CONFIG_X86_64
+#undef CONFIG_COMPAT
 #undef CONFIG_PGTABLE_LEVELS
 #undef CONFIG_ILLEGAL_POINTER_VALUE
 #undef CONFIG_SPARSEMEM_VMEMMAP

arch/x86/ia32/Makefile

@@ -3,7 +3,7 @@
 # Makefile for the ia32 kernel emulation subsystem.
 #
 
-obj-$(CONFIG_IA32_EMULATION) := sys_ia32.o ia32_signal.o
+obj-$(CONFIG_IA32_EMULATION) := ia32_signal.o
 
 obj-$(CONFIG_IA32_AOUT) += ia32_aout.o
 

arch/x86/include/asm/irq.h

@@ -36,7 +36,7 @@ extern void native_init_IRQ(void);
 
 extern void handle_irq(struct irq_desc *desc, struct pt_regs *regs);
 
-extern __visible unsigned int do_IRQ(struct pt_regs *regs);
+extern __visible void do_IRQ(struct pt_regs *regs);
 
 extern void init_ISA_irqs(void);
 

arch/x86/include/asm/mce.h

@@ -238,9 +238,6 @@ extern void mce_disable_bank(int bank);
 /*
  * Exception handler
  */
-
-/* Call the installed machine check handler for this CPU setup. */
-extern void (*machine_check_vector)(struct pt_regs *, long error_code);
 void do_machine_check(struct pt_regs *, long);
 
 /*

arch/x86/include/asm/preempt.h

@@ -103,14 +103,14 @@ static __always_inline bool should_resched(int preempt_offset)
 }
 
 #ifdef CONFIG_PREEMPTION
-  extern asmlinkage void ___preempt_schedule(void);
+  extern asmlinkage void preempt_schedule_thunk(void);
 # define __preempt_schedule() \
-	asm volatile ("call ___preempt_schedule" : ASM_CALL_CONSTRAINT)
+	asm volatile ("call preempt_schedule_thunk" : ASM_CALL_CONSTRAINT)
 
   extern asmlinkage void preempt_schedule(void);
-  extern asmlinkage void ___preempt_schedule_notrace(void);
+  extern asmlinkage void preempt_schedule_notrace_thunk(void);
 # define __preempt_schedule_notrace() \
-	asm volatile ("call ___preempt_schedule_notrace" : ASM_CALL_CONSTRAINT)
+	asm volatile ("call preempt_schedule_notrace_thunk" : ASM_CALL_CONSTRAINT)
 
   extern asmlinkage void preempt_schedule_notrace(void);
 #endif

arch/x86/include/asm/sighandling.h

@@ -17,9 +17,4 @@ void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
 int setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate,
 		     struct pt_regs *regs, unsigned long mask);
 
-
-#ifdef CONFIG_X86_X32_ABI
-asmlinkage long sys32_x32_rt_sigreturn(void);
-#endif
-
 #endif /* _ASM_X86_SIGHANDLING_H */