x86/entry: Use PUSH_AND_CLEAR_GPRS and drop SAVE_ALL

PUSH_AND_CLEAR_GPRS is shorter than SAVE_ALL in terms of emitted code.

lstar_enter() and cstar_enter() do not need CLD.  The SYSCALL instruction
sanitises flags based on MSR_SYSCALL_MASK.  For all other cases, place the CLD
instruction next to STAC/CLAC so the flag handling is together.

Get rid of the compat=1 special case for PV32.  It's not obviously a win, and
PV32 is getting increasingly rare these days.

No functional change.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

xen/arch/x86/include/asm/asm_defns.h

@@ -219,51 +219,6 @@ static always_inline void stac(void)
 #endif
 
 #ifdef __ASSEMBLER__
-.macro SAVE_ALL compat=0
-        addq  $-(UREGS_error_code-UREGS_r15), %rsp
-        cld
-        movq  %rdi,UREGS_rdi(%rsp)
-        xor   %edi, %edi
-        movq  %rsi,UREGS_rsi(%rsp)
-        xor   %esi, %esi
-        movq  %rdx,UREGS_rdx(%rsp)
-        xor   %edx, %edx
-        movq  %rcx,UREGS_rcx(%rsp)
-        xor   %ecx, %ecx
-        movq  %rax,UREGS_rax(%rsp)
-        xor   %eax, %eax
-.if !\compat
-        movq  %r8,UREGS_r8(%rsp)
-        movq  %r9,UREGS_r9(%rsp)
-        movq  %r10,UREGS_r10(%rsp)
-        movq  %r11,UREGS_r11(%rsp)
-.endif
-        xor   %r8d, %r8d
-        xor   %r9d, %r9d
-        xor   %r10d, %r10d
-        xor   %r11d, %r11d
-        movq  %rbx,UREGS_rbx(%rsp)
-        xor   %ebx, %ebx
-        movq  %rbp,UREGS_rbp(%rsp)
-#ifdef CONFIG_FRAME_POINTER
-/* Indicate special exception stack frame by inverting the frame pointer. */
-        leaq  UREGS_rbp(%rsp), %rbp
-        notq  %rbp
-#else
-        xor   %ebp, %ebp
-#endif
-.if !\compat
-        movq  %r12,UREGS_r12(%rsp)
-        movq  %r13,UREGS_r13(%rsp)
-        movq  %r14,UREGS_r14(%rsp)
-        movq  %r15,UREGS_r15(%rsp)
-.endif
-        xor   %r12d, %r12d
-        xor   %r13d, %r13d
-        xor   %r14d, %r14d
-        xor   %r15d, %r15d
-.endm
-
 /*
  * Push and clear GPRs
  */

xen/arch/x86/x86_64/compat/entry.S

@@ -11,12 +11,14 @@
 #include <asm/desc.h>
 #include <public/xen.h>
 
+/* DPL1 gate, restricted to 32bit PV guest kernels only. */
 FUNC(entry_int82)
         ENDBR64
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
         pushq $0
         movb  $HYPERCALL_VECTOR, EFRAME_entry_vector(%rsp)
-        SAVE_ALL compat=1 /* DPL1 gate, restricted to 32bit PV guests only. */
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 

xen/arch/x86/x86_64/entry.S

@@ -305,7 +305,7 @@ FUNC(lstar_enter)
         pushq $0
         BUILD_BUG_ON(TRAP_syscall & 0xff)
         movb  $TRAP_syscall >> 8, EFRAME_entry_vector + 1(%rsp)
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -345,7 +345,7 @@ FUNC(cstar_enter)
         pushq $0
         BUILD_BUG_ON(TRAP_syscall & 0xff)
         movb  $TRAP_syscall >> 8, EFRAME_entry_vector + 1(%rsp)
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -383,13 +383,14 @@ FUNC(sysenter_entry)
         pushq $0
         pushfq
 LABEL(sysenter_eflags_saved, 0)
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
         pushq $3 /* ring 3 null cs */
         pushq $0 /* null rip */
         pushq $0
         BUILD_BUG_ON(TRAP_syscall & 0xff)
         movb  $TRAP_syscall >> 8, EFRAME_entry_vector + 1(%rsp)
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -442,10 +443,11 @@ END(sysenter_entry)
 
 FUNC(entry_int80)
         ENDBR64
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
         pushq $0
         movb  $0x80, EFRAME_entry_vector(%rsp)
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -722,8 +724,9 @@ END(ret_from_intr)
         .section .init.text, "ax", @progbits
 FUNC(early_page_fault)
         ENDBR64
+        cld
         movb  $X86_EXC_PF, EFRAME_entry_vector(%rsp)
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
         movq  %rsp, %rdi
         call  do_early_page_fault
         jmp   restore_all_xen
@@ -769,8 +772,9 @@ UNLIKELY_END(exit_cr3)
 END(restore_all_xen)
 
 FUNC(common_interrupt)
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -805,8 +809,9 @@ FUNC(entry_PF)
 END(entry_PF)
 /* No special register assumptions. */
 FUNC(handle_exception, 0)
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -1103,9 +1108,10 @@ END(entry_CP)
 FUNC(entry_DF)
         ENDBR64
         movb  $X86_EXC_DF, EFRAME_entry_vector(%rsp)
+        cld
         /* Set AC to reduce chance of further SMAP faults */
         ALTERNATIVE "", stac, X86_FEATURE_XEN_SMAP
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)
 
@@ -1132,8 +1138,9 @@ FUNC(entry_NMI)
 END(entry_NMI)
 
 FUNC(handle_ist_exception)
+        cld
         ALTERNATIVE "", clac, X86_FEATURE_XEN_SMAP
-        SAVE_ALL
+        PUSH_AND_CLEAR_GPRS
 
         GET_STACK_END(14)