[PATCH] x86/entry/64: Remove duplicate syscall table for fast path
From: Brian Gerst
Date: Wed Dec 09 2015 - 08:03:01 EST
Instead of using a duplicate syscall table for the fast path, create stubs for
the syscalls that need pt_regs that save the extra registers if a flag for the
slow path is not set.
Signed-off-by: Brian Gerst <brgerst@xxxxxxxxx>
To: Andy Lutomirski <luto@xxxxxxxxxxxxxx>
Cc: Andy Lutomirski <luto@xxxxxxxxxx>
Cc: the arch/x86 maintainers <x86@xxxxxxxxxx>
Cc: Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: FrÃdÃric Weisbecker <fweisbec@xxxxxxxxx>
Cc: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
---
Applies on top of Andy's syscall cleanup series.
arch/x86/entry/calling.h | 32 ++++++++++++++++----------------
arch/x86/entry/common.c | 4 ++++
arch/x86/entry/entry_64.S | 36 +++++++++++++++++++++++++++++-------
arch/x86/entry/syscall_64.c | 25 +++++--------------------
arch/x86/include/asm/thread_info.h | 1 +
5 files changed, 55 insertions(+), 43 deletions(-)
diff --git a/arch/x86/entry/calling.h b/arch/x86/entry/calling.h
index e32206e..7c58bd2 100644
--- a/arch/x86/entry/calling.h
+++ b/arch/x86/entry/calling.h
@@ -129,22 +129,22 @@ For 32-bit we have the following conventions - kernel is built with
SAVE_C_REGS_HELPER 0, 0, 0, 1, 0
.endm
- .macro SAVE_EXTRA_REGS offset=0
- movq %r15, 0*8+\offset(%rsp)
- movq %r14, 1*8+\offset(%rsp)
- movq %r13, 2*8+\offset(%rsp)
- movq %r12, 3*8+\offset(%rsp)
- movq %rbp, 4*8+\offset(%rsp)
- movq %rbx, 5*8+\offset(%rsp)
- .endm
-
- .macro RESTORE_EXTRA_REGS offset=0
- movq 0*8+\offset(%rsp), %r15
- movq 1*8+\offset(%rsp), %r14
- movq 2*8+\offset(%rsp), %r13
- movq 3*8+\offset(%rsp), %r12
- movq 4*8+\offset(%rsp), %rbp
- movq 5*8+\offset(%rsp), %rbx
+ .macro SAVE_EXTRA_REGS offset=0 base=rsp
+ movq %r15, 0*8+\offset(%\base)
+ movq %r14, 1*8+\offset(%\base)
+ movq %r13, 2*8+\offset(%\base)
+ movq %r12, 3*8+\offset(%\base)
+ movq %rbp, 4*8+\offset(%\base)
+ movq %rbx, 5*8+\offset(%\base)
+ .endm
+
+ .macro RESTORE_EXTRA_REGS offset=0 base=rsp
+ movq 0*8+\offset(%\base), %r15
+ movq 1*8+\offset(%\base), %r14
+ movq 2*8+\offset(%\base), %r13
+ movq 3*8+\offset(%\base), %r12
+ movq 4*8+\offset(%\base), %rbp
+ movq 5*8+\offset(%\base), %rbx
.endm
.macro ZERO_EXTRA_REGS
diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c
index 016ac47..4381aca 100644
--- a/arch/x86/entry/common.c
+++ b/arch/x86/entry/common.c
@@ -342,6 +342,8 @@ __visible void do_syscall_64(struct pt_regs *regs)
struct thread_info *ti = pt_regs_to_thread_info(regs);
unsigned long nr = regs->orig_ax;
+ ti->status |= TS_SLOWPATH;
+
enter_from_user_mode();
local_irq_enable();
@@ -360,6 +362,8 @@ __visible void do_syscall_64(struct pt_regs *regs)
}
syscall_return_slowpath(regs);
+
+ ti->status &= ~TS_SLOWPATH;
}
#endif
diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
index 1ab5362..5852ec6 100644
--- a/arch/x86/entry/entry_64.S
+++ b/arch/x86/entry/entry_64.S
@@ -188,7 +188,7 @@ entry_SYSCALL_64_fastpath:
#endif
ja 1f /* return -ENOSYS (already in pt_regs->ax) */
movq %r10, %rcx
- call *sys_call_table_fastpath_64(, %rax, 8)
+ call *sys_call_table(, %rax, 8)
movq %rax, RAX(%rsp)
1:
@@ -306,15 +306,37 @@ END(entry_SYSCALL_64)
ENTRY(stub_ptregs_64)
/*
- * Syscalls marked as needing ptregs that go through the fast path
- * land here. We transfer to the slow path.
+ * Syscalls marked as needing ptregs land here.
+ * If we are on the fast path, we need to save the extra regs.
+ * If we are on the slow path, the extra regs are already saved.
*/
- DISABLE_INTERRUPTS(CLBR_NONE)
- TRACE_IRQS_OFF
- addq $8, %rsp
- jmp entry_SYSCALL64_slow_path
+ movq PER_CPU_VAR(cpu_current_top_of_stack), %r10
+ testl $TS_SLOWPATH, ASM_THREAD_INFO(TI_status, %r10, 0)
+ jnz 1f
+ subq $SIZEOF_PTREGS, %r10
+ SAVE_EXTRA_REGS base=r10
+ movq %r10, %rbx
+ call *%rax
+ movq %rbx, %r10
+ RESTORE_EXTRA_REGS base=r10
+ ret
+1:
+ jmp *%rax
END(stub_ptregs_64)
+.macro ptregs_stub func
+ENTRY(ptregs_\func)
+ leaq \func(%rip), %rax
+ jmp stub_ptregs_64
+END(ptregs_\func)
+.endm
+
+#define __SYSCALL_64_QUAL_(sym)
+#define __SYSCALL_64_QUAL_ptregs(sym) ptregs_stub sym
+
+#define __SYSCALL_64(nr, sym, qual) __SYSCALL_64_QUAL_##qual(sym)
+#include <asm/syscalls_64.h>
+
/*
* A newly forked process directly context switches into this address.
*
diff --git a/arch/x86/entry/syscall_64.c b/arch/x86/entry/syscall_64.c
index 601745c..9dbc5ab 100644
--- a/arch/x86/entry/syscall_64.c
+++ b/arch/x86/entry/syscall_64.c
@@ -6,11 +6,14 @@
#include <asm/asm-offsets.h>
#include <asm/syscall.h>
-#define __SYSCALL_64(nr, sym, qual) extern asmlinkage long sym(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long) ;
+#define __SYSCALL_64_QUAL_(sym) sym
+#define __SYSCALL_64_QUAL_ptregs(sym) ptregs_##sym
+
+#define __SYSCALL_64(nr, sym, qual) extern asmlinkage long __SYSCALL_64_QUAL_##qual(sym)(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
#include <asm/syscalls_64.h>
#undef __SYSCALL_64
-#define __SYSCALL_64(nr, sym, qual) [nr] = sym,
+#define __SYSCALL_64(nr, sym, qual) [nr] = __SYSCALL_64_QUAL_##qual(sym),
extern long sys_ni_syscall(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
@@ -22,21 +25,3 @@ asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
[0 ... __NR_syscall_max] = &sys_ni_syscall,
#include <asm/syscalls_64.h>
};
-
-#undef __SYSCALL_64
-
-extern long stub_ptregs_64(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
-
-#define __SYSCALL_64_QUAL_(nr, sym) [nr] = sym,
-#define __SYSCALL_64_QUAL_ptregs(nr, sym) [nr] = stub_ptregs_64,
-
-#define __SYSCALL_64(nr, sym, qual) __SYSCALL_64_QUAL_##qual(nr, sym)
-
-asmlinkage const sys_call_ptr_t sys_call_table_fastpath_64[__NR_syscall_max+1] = {
- /*
- * Smells like a compiler bug -- it doesn't work
- * when the & below is removed.
- */
- [0 ... __NR_syscall_max] = &sys_ni_syscall,
-#include <asm/syscalls_64.h>
-};
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index ae210d6..358e3a9 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -229,6 +229,7 @@ static inline unsigned long current_stack_pointer(void)
* ever touches our thread-synchronous status, so we don't
* have to worry about atomic accesses.
*/
+#define TS_SLOWPATH 0x0001 /* syscall slowpath (64BIT) */
#define TS_COMPAT 0x0002 /* 32bit syscall active (64BIT)*/
#define TS_RESTORE_SIGMASK 0x0008 /* restore signal mask in do_signal() */
--
2.5.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/