[PATCH v4 00/21] paravirt: cleanup and reorg

[PATCH v4 00/21] paravirt: cleanup and reorg

[Juergen Gross]

Some cleanups and reorg of paravirt code and headers:

- The first 2 patches should be not controversial at all, as they
  remove just some no longer needed #include and struct forward
  declarations.

- The 3rd patch is removing CONFIG_PARAVIRT_DEBUG, which IMO has
  no real value, as it just changes a crash to a BUG() (the stack
  trace will basically be the same). As the maintainer of the main
  paravirt user (Xen) I have never seen this crash/BUG() to happen.

- The 4th patch is just a movement of code.

- I don't know for what reason asm/paravirt_api_clock.h was added,
  as all archs supporting it do it exactly in the same way. Patch
  5 is removing it.

- Patches 6-14 are streamlining the paravirt clock interfaces by
  using a common implementation across architectures where possible
  and by moving the related code into common sched code, as this is
  where it should live.

- Patches 15-20 are more like RFC material preparing the paravirt
  infrastructure to support multiple pv_ops function arrays.
  As a prerequisite for that it makes life in objtool much easier
  with dropping the Xen static initializers of the pv_ops sub-
  structures, which is done in patches 15-17.
  Patches 18-20 are doing the real preparations for multiple pv_ops
  arrays and using those arrays in multiple headers.

- Patch 21 is an example how the new scheme can look like using the
  PV-spinlocks.

Changes in V2:
- new patches 13-18 and 20
- complete rework of patch 21

Changes in V3:
- fixed 2 issues detected by kernel test robot

Changes in V4:
- fixed one build issue

Juergen Gross (21):
  x86/paravirt: Remove not needed includes of paravirt.h
  x86/paravirt: Remove some unneeded struct declarations
  x86/paravirt: Remove PARAVIRT_DEBUG config option
  x86/paravirt: Move thunk macros to paravirt_types.h
  paravirt: Remove asm/paravirt_api_clock.h
  sched: Move clock related paravirt code to kernel/sched
  arm/paravirt: Use common code for paravirt_steal_clock()
  arm64/paravirt: Use common code for paravirt_steal_clock()
  loongarch/paravirt: Use common code for paravirt_steal_clock()
  riscv/paravirt: Use common code for paravirt_steal_clock()
  x86/paravirt: Use common code for paravirt_steal_clock()
  x86/paravirt: Move paravirt_sched_clock() related code into tsc.c
  x86/paravirt: Introduce new paravirt-base.h header
  x86/paravirt: Move pv_native_*() prototypes to paravirt.c
  x86/xen: Drop xen_irq_ops
  x86/xen: Drop xen_cpu_ops
  x86/xen: Drop xen_mmu_ops
  objtool: Allow multiple pv_ops arrays
  x86/paravirt: Allow pv-calls outside paravirt.h
  x86/paravirt: Specify pv_ops array in paravirt macros
  x86/pvlocks: Move paravirt spinlock functions into own header

 arch/Kconfig                                  |   3 +
 arch/arm/Kconfig                              |   1 +
 arch/arm/include/asm/paravirt.h               |  22 --
 arch/arm/include/asm/paravirt_api_clock.h     |   1 -
 arch/arm/kernel/Makefile                      |   1 -
 arch/arm/kernel/paravirt.c                    |  23 --
 arch/arm64/Kconfig                            |   1 +
 arch/arm64/include/asm/paravirt.h             |  14 -
 arch/arm64/include/asm/paravirt_api_clock.h   |   1 -
 arch/arm64/kernel/paravirt.c                  |  11 +-
 arch/loongarch/Kconfig                        |   1 +
 arch/loongarch/include/asm/paravirt.h         |  13 -
 .../include/asm/paravirt_api_clock.h          |   1 -
 arch/loongarch/kernel/paravirt.c              |  10 +-
 arch/powerpc/include/asm/paravirt.h           |   3 -
 arch/powerpc/include/asm/paravirt_api_clock.h |   2 -
 arch/powerpc/platforms/pseries/setup.c        |   4 +-
 arch/riscv/Kconfig                            |   1 +
 arch/riscv/include/asm/paravirt.h             |  14 -
 arch/riscv/include/asm/paravirt_api_clock.h   |   1 -
 arch/riscv/kernel/paravirt.c                  |  11 +-
 arch/x86/Kconfig                              |   8 +-
 arch/x86/entry/entry_64.S                     |   1 -
 arch/x86/entry/vsyscall/vsyscall_64.c         |   1 -
 arch/x86/hyperv/hv_spinlock.c                 |  11 +-
 arch/x86/include/asm/apic.h                   |   4 -
 arch/x86/include/asm/highmem.h                |   1 -
 arch/x86/include/asm/mshyperv.h               |   1 -
 arch/x86/include/asm/paravirt-base.h          |  29 ++
 arch/x86/include/asm/paravirt-spinlock.h      | 146 ++++++++
 arch/x86/include/asm/paravirt.h               | 331 +++++-------------
 arch/x86/include/asm/paravirt_api_clock.h     |   1 -
 arch/x86/include/asm/paravirt_types.h         | 269 +++++++-------
 arch/x86/include/asm/pgtable_32.h             |   1 -
 arch/x86/include/asm/ptrace.h                 |   2 +-
 arch/x86/include/asm/qspinlock.h              |  89 +----
 arch/x86/include/asm/spinlock.h               |   1 -
 arch/x86/include/asm/timer.h                  |   1 +
 arch/x86/include/asm/tlbflush.h               |   4 -
 arch/x86/kernel/Makefile                      |   2 +-
 arch/x86/kernel/apm_32.c                      |   1 -
 arch/x86/kernel/callthunks.c                  |   1 -
 arch/x86/kernel/cpu/bugs.c                    |   1 -
 arch/x86/kernel/cpu/vmware.c                  |   1 +
 arch/x86/kernel/kvm.c                         |  13 +-
 arch/x86/kernel/kvmclock.c                    |   1 +
 arch/x86/kernel/paravirt-spinlocks.c          |  26 +-
 arch/x86/kernel/paravirt.c                    |  42 +--
 arch/x86/kernel/tsc.c                         |  10 +-
 arch/x86/kernel/vsmp_64.c                     |   1 -
 arch/x86/lib/cache-smp.c                      |   1 -
 arch/x86/mm/init.c                            |   1 -
 arch/x86/xen/enlighten_pv.c                   |  82 ++---
 arch/x86/xen/irq.c                            |  20 +-
 arch/x86/xen/mmu_pv.c                         | 100 ++----
 arch/x86/xen/spinlock.c                       |  11 +-
 arch/x86/xen/time.c                           |   2 +
 drivers/clocksource/hyperv_timer.c            |   2 +
 drivers/xen/time.c                            |   2 +-
 include/linux/sched/cputime.h                 |  18 +
 kernel/sched/core.c                           |   5 +
 kernel/sched/cputime.c                        |  13 +
 kernel/sched/sched.h                          |   3 +-
 tools/objtool/arch/x86/decode.c               |   8 +-
 tools/objtool/check.c                         |  78 ++++-
 tools/objtool/include/objtool/check.h         |   2 +
 66 files changed, 661 insertions(+), 826 deletions(-)
 delete mode 100644 arch/arm/include/asm/paravirt.h
 delete mode 100644 arch/arm/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/arm/kernel/paravirt.c
 delete mode 100644 arch/arm64/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/loongarch/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/powerpc/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/riscv/include/asm/paravirt_api_clock.h
 create mode 100644 arch/x86/include/asm/paravirt-base.h
 create mode 100644 arch/x86/include/asm/paravirt-spinlock.h
 delete mode 100644 arch/x86/include/asm/paravirt_api_clock.h

-- 
2.51.0

[PATCH v4 02/21] x86/paravirt: Remove some unneeded struct declarations

[Juergen Gross]

In paravirt_types.h and paravirt.h there are some struct declarations
which are not needed. Remove them.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
V2:
- remove mm_struct from paravirt.h, too
---
 arch/x86/include/asm/paravirt.h       | 4 ----
 arch/x86/include/asm/paravirt_types.h | 6 ------
 2 files changed, 10 deletions(-)

diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index b5e59a7ba0d0..612b3df65b1b 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -6,10 +6,6 @@
 
 #include <asm/paravirt_types.h>
 
-#ifndef __ASSEMBLER__
-struct mm_struct;
-#endif
-
 #ifdef CONFIG_PARAVIRT
 #include <asm/pgtable_types.h>
 #include <asm/asm.h>
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 37a8627d8277..84cc8c95713b 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -11,16 +11,11 @@
 #include <asm/pgtable_types.h>
 #include <asm/nospec-branch.h>
 
-struct page;
 struct thread_struct;
-struct desc_ptr;
-struct tss_struct;
 struct mm_struct;
-struct desc_struct;
 struct task_struct;
 struct cpumask;
 struct flush_tlb_info;
-struct mmu_gather;
 struct vm_area_struct;
 
 /*
@@ -205,7 +200,6 @@ struct pv_mmu_ops {
 #endif
 } __no_randomize_layout;
 
-struct arch_spinlock;
 #ifdef CONFIG_SMP
 #include <asm/spinlock_types.h>
 #endif
-- 
2.51.0

[PATCH v4 03/21] x86/paravirt: Remove PARAVIRT_DEBUG config option

[Juergen Gross]

The only effect of CONFIG_PARAVIRT_DEBUG set is that instead of doing
a call using a NULL pointer a BUG() is being raised.

While the BUG() will be a little bit easier to analyse, the call of
NULL isn't really that difficult to find the reason for.

Remove the config option to make paravirt coding a little bit less
annoying.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/x86/Kconfig                      | 7 -------
 arch/x86/include/asm/paravirt.h       | 1 -
 arch/x86/include/asm/paravirt_types.h | 8 --------
 3 files changed, 16 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index fa3b616af03a..aa279fec5c1f 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -805,13 +805,6 @@ config PARAVIRT_XXL
 	bool
 	depends on X86_64
 
-config PARAVIRT_DEBUG
-	bool "paravirt-ops debugging"
-	depends on PARAVIRT && DEBUG_KERNEL
-	help
-	  Enable to debug paravirt_ops internals.  Specifically, BUG if
-	  a paravirt_op is missing when it is called.
-
 config PARAVIRT_SPINLOCKS
 	bool "Paravirtualization layer for spinlocks"
 	depends on PARAVIRT && SMP
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index 612b3df65b1b..fd9826397419 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -12,7 +12,6 @@
 #include <asm/nospec-branch.h>
 
 #ifndef __ASSEMBLER__
-#include <linux/bug.h>
 #include <linux/types.h>
 #include <linux/cpumask.h>
 #include <linux/static_call_types.h>
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 84cc8c95713b..085095f94f97 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -354,12 +354,6 @@ extern struct paravirt_patch_template pv_ops;
 #define VEXTRA_CLOBBERS	 , "rax", "r8", "r9", "r10", "r11"
 #endif	/* CONFIG_X86_32 */
 
-#ifdef CONFIG_PARAVIRT_DEBUG
-#define PVOP_TEST_NULL(op)	BUG_ON(pv_ops.op == NULL)
-#else
-#define PVOP_TEST_NULL(op)	((void)pv_ops.op)
-#endif
-
 #define PVOP_RETVAL(rettype)						\
 	({	unsigned long __mask = ~0UL;				\
 		BUILD_BUG_ON(sizeof(rettype) > sizeof(unsigned long));	\
@@ -388,7 +382,6 @@ extern struct paravirt_patch_template pv_ops;
 #define ____PVOP_CALL(ret, op, call_clbr, extra_clbr, ...)	\
 	({								\
 		PVOP_CALL_ARGS;						\
-		PVOP_TEST_NULL(op);					\
 		asm volatile(ALTERNATIVE(PARAVIRT_CALL, ALT_CALL_INSTR,	\
 				ALT_CALL_ALWAYS)			\
 			     : call_clbr, ASM_CALL_CONSTRAINT		\
@@ -402,7 +395,6 @@ extern struct paravirt_patch_template pv_ops;
 			  extra_clbr, ...)				\
 	({								\
 		PVOP_CALL_ARGS;						\
-		PVOP_TEST_NULL(op);					\
 		asm volatile(ALTERNATIVE_2(PARAVIRT_CALL,		\
 				 ALT_CALL_INSTR, ALT_CALL_ALWAYS,	\
 				 alt, cond)				\
-- 
2.51.0

[PATCH v4 04/21] x86/paravirt: Move thunk macros to paravirt_types.h

[Juergen Gross]

The macros for generating PV-thunks are part of the generic paravirt
infrastructure, so they should be in paravirt_types.h.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/x86/include/asm/paravirt.h       | 68 ---------------------------
 arch/x86/include/asm/paravirt_types.h | 68 +++++++++++++++++++++++++++
 2 files changed, 68 insertions(+), 68 deletions(-)

diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index fd9826397419..1344d2fb2b86 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -581,74 +581,6 @@ bool __raw_callee_save___native_vcpu_is_preempted(long cpu);
 
 #endif /* SMP && PARAVIRT_SPINLOCKS */
 
-#ifdef CONFIG_X86_32
-/* save and restore all caller-save registers, except return value */
-#define PV_SAVE_ALL_CALLER_REGS		"pushl %ecx;"
-#define PV_RESTORE_ALL_CALLER_REGS	"popl  %ecx;"
-#else
-/* save and restore all caller-save registers, except return value */
-#define PV_SAVE_ALL_CALLER_REGS						\
-	"push %rcx;"							\
-	"push %rdx;"							\
-	"push %rsi;"							\
-	"push %rdi;"							\
-	"push %r8;"							\
-	"push %r9;"							\
-	"push %r10;"							\
-	"push %r11;"
-#define PV_RESTORE_ALL_CALLER_REGS					\
-	"pop %r11;"							\
-	"pop %r10;"							\
-	"pop %r9;"							\
-	"pop %r8;"							\
-	"pop %rdi;"							\
-	"pop %rsi;"							\
-	"pop %rdx;"							\
-	"pop %rcx;"
-#endif
-
-/*
- * Generate a thunk around a function which saves all caller-save
- * registers except for the return value.  This allows C functions to
- * be called from assembler code where fewer than normal registers are
- * available.  It may also help code generation around calls from C
- * code if the common case doesn't use many registers.
- *
- * When a callee is wrapped in a thunk, the caller can assume that all
- * arg regs and all scratch registers are preserved across the
- * call. The return value in rax/eax will not be saved, even for void
- * functions.
- */
-#define PV_THUNK_NAME(func) "__raw_callee_save_" #func
-#define __PV_CALLEE_SAVE_REGS_THUNK(func, section)			\
-	extern typeof(func) __raw_callee_save_##func;			\
-									\
-	asm(".pushsection " section ", \"ax\";"				\
-	    ".globl " PV_THUNK_NAME(func) ";"				\
-	    ".type " PV_THUNK_NAME(func) ", @function;"			\
-	    ASM_FUNC_ALIGN						\
-	    PV_THUNK_NAME(func) ":"					\
-	    ASM_ENDBR							\
-	    FRAME_BEGIN							\
-	    PV_SAVE_ALL_CALLER_REGS					\
-	    "call " #func ";"						\
-	    PV_RESTORE_ALL_CALLER_REGS					\
-	    FRAME_END							\
-	    ASM_RET							\
-	    ".size " PV_THUNK_NAME(func) ", .-" PV_THUNK_NAME(func) ";"	\
-	    ".popsection")
-
-#define PV_CALLEE_SAVE_REGS_THUNK(func)			\
-	__PV_CALLEE_SAVE_REGS_THUNK(func, ".text")
-
-/* Get a reference to a callee-save function */
-#define PV_CALLEE_SAVE(func)						\
-	((struct paravirt_callee_save) { __raw_callee_save_##func })
-
-/* Promise that "func" already uses the right calling convention */
-#define __PV_IS_CALLEE_SAVE(func)			\
-	((struct paravirt_callee_save) { func })
-
 #ifdef CONFIG_PARAVIRT_XXL
 static __always_inline unsigned long arch_local_save_flags(void)
 {
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 085095f94f97..7acff40cc159 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -512,5 +512,73 @@ unsigned long pv_native_read_cr2(void);
 
 #define ALT_NOT_XEN	ALT_NOT(X86_FEATURE_XENPV)
 
+#ifdef CONFIG_X86_32
+/* save and restore all caller-save registers, except return value */
+#define PV_SAVE_ALL_CALLER_REGS		"pushl %ecx;"
+#define PV_RESTORE_ALL_CALLER_REGS	"popl  %ecx;"
+#else
+/* save and restore all caller-save registers, except return value */
+#define PV_SAVE_ALL_CALLER_REGS						\
+	"push %rcx;"							\
+	"push %rdx;"							\
+	"push %rsi;"							\
+	"push %rdi;"							\
+	"push %r8;"							\
+	"push %r9;"							\
+	"push %r10;"							\
+	"push %r11;"
+#define PV_RESTORE_ALL_CALLER_REGS					\
+	"pop %r11;"							\
+	"pop %r10;"							\
+	"pop %r9;"							\
+	"pop %r8;"							\
+	"pop %rdi;"							\
+	"pop %rsi;"							\
+	"pop %rdx;"							\
+	"pop %rcx;"
+#endif
+
+/*
+ * Generate a thunk around a function which saves all caller-save
+ * registers except for the return value.  This allows C functions to
+ * be called from assembler code where fewer than normal registers are
+ * available.  It may also help code generation around calls from C
+ * code if the common case doesn't use many registers.
+ *
+ * When a callee is wrapped in a thunk, the caller can assume that all
+ * arg regs and all scratch registers are preserved across the
+ * call. The return value in rax/eax will not be saved, even for void
+ * functions.
+ */
+#define PV_THUNK_NAME(func) "__raw_callee_save_" #func
+#define __PV_CALLEE_SAVE_REGS_THUNK(func, section)			\
+	extern typeof(func) __raw_callee_save_##func;			\
+									\
+	asm(".pushsection " section ", \"ax\";"				\
+	    ".globl " PV_THUNK_NAME(func) ";"				\
+	    ".type " PV_THUNK_NAME(func) ", @function;"			\
+	    ASM_FUNC_ALIGN						\
+	    PV_THUNK_NAME(func) ":"					\
+	    ASM_ENDBR							\
+	    FRAME_BEGIN							\
+	    PV_SAVE_ALL_CALLER_REGS					\
+	    "call " #func ";"						\
+	    PV_RESTORE_ALL_CALLER_REGS					\
+	    FRAME_END							\
+	    ASM_RET							\
+	    ".size " PV_THUNK_NAME(func) ", .-" PV_THUNK_NAME(func) ";"	\
+	    ".popsection")
+
+#define PV_CALLEE_SAVE_REGS_THUNK(func)			\
+	__PV_CALLEE_SAVE_REGS_THUNK(func, ".text")
+
+/* Get a reference to a callee-save function */
+#define PV_CALLEE_SAVE(func)						\
+	((struct paravirt_callee_save) { __raw_callee_save_##func })
+
+/* Promise that "func" already uses the right calling convention */
+#define __PV_IS_CALLEE_SAVE(func)			\
+	((struct paravirt_callee_save) { func })
+
 #endif  /* CONFIG_PARAVIRT */
 #endif	/* _ASM_X86_PARAVIRT_TYPES_H */
-- 
2.51.0

[PATCH v4 05/21] paravirt: Remove asm/paravirt_api_clock.h

[Juergen Gross]

All architectures supporting CONFIG_PARAVIRT share the same contents
of asm/paravirt_api_clock.h:

  #include <asm/paravirt.h>

So remove all incarnations of asm/paravirt_api_clock.h and remove the
only place where it is included, as there asm/paravirt.h is included
anyway.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com> # powerpc, scheduler bits
---
 arch/arm/include/asm/paravirt_api_clock.h       | 1 -
 arch/arm64/include/asm/paravirt_api_clock.h     | 1 -
 arch/loongarch/include/asm/paravirt_api_clock.h | 1 -
 arch/powerpc/include/asm/paravirt_api_clock.h   | 2 --
 arch/riscv/include/asm/paravirt_api_clock.h     | 1 -
 arch/x86/include/asm/paravirt_api_clock.h       | 1 -
 kernel/sched/sched.h                            | 1 -
 7 files changed, 8 deletions(-)
 delete mode 100644 arch/arm/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/arm64/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/loongarch/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/powerpc/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/riscv/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/x86/include/asm/paravirt_api_clock.h

diff --git a/arch/arm/include/asm/paravirt_api_clock.h b/arch/arm/include/asm/paravirt_api_clock.h
deleted file mode 100644
index 65ac7cee0dad..000000000000
--- a/arch/arm/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1 +0,0 @@
-#include <asm/paravirt.h>
diff --git a/arch/arm64/include/asm/paravirt_api_clock.h b/arch/arm64/include/asm/paravirt_api_clock.h
deleted file mode 100644
index 65ac7cee0dad..000000000000
--- a/arch/arm64/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1 +0,0 @@
-#include <asm/paravirt.h>
diff --git a/arch/loongarch/include/asm/paravirt_api_clock.h b/arch/loongarch/include/asm/paravirt_api_clock.h
deleted file mode 100644
index 65ac7cee0dad..000000000000
--- a/arch/loongarch/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1 +0,0 @@
-#include <asm/paravirt.h>
diff --git a/arch/powerpc/include/asm/paravirt_api_clock.h b/arch/powerpc/include/asm/paravirt_api_clock.h
deleted file mode 100644
index d25ca7ac57c7..000000000000
--- a/arch/powerpc/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1,2 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#include <asm/paravirt.h>
diff --git a/arch/riscv/include/asm/paravirt_api_clock.h b/arch/riscv/include/asm/paravirt_api_clock.h
deleted file mode 100644
index 65ac7cee0dad..000000000000
--- a/arch/riscv/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1 +0,0 @@
-#include <asm/paravirt.h>
diff --git a/arch/x86/include/asm/paravirt_api_clock.h b/arch/x86/include/asm/paravirt_api_clock.h
deleted file mode 100644
index 65ac7cee0dad..000000000000
--- a/arch/x86/include/asm/paravirt_api_clock.h
+++ /dev/null
@@ -1 +0,0 @@
-#include <asm/paravirt.h>
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index adfb6e3409d7..ccd0d92ff7d1 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -84,7 +84,6 @@ struct cpuidle_state;
 
 #ifdef CONFIG_PARAVIRT
 # include <asm/paravirt.h>
-# include <asm/paravirt_api_clock.h>
 #endif
 
 #include <asm/barrier.h>
-- 
2.51.0

[PATCH v4 06/21] sched: Move clock related paravirt code to kernel/sched

[Juergen Gross]

Paravirt clock related functions are available in multiple archs.

In order to share the common parts, move the common static keys
to kernel/sched/ and remove them from the arch specific files.

Make a common paravirt_steal_clock() implementation available in
kernel/sched/cputime.c, guarding it with a new config option
CONFIG_HAVE_PV_STEAL_CLOCK_GEN, which can be selected by an arch
in case it wants to use that common variant.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/Kconfig                           |  3 +++
 arch/arm/include/asm/paravirt.h        |  4 ----
 arch/arm/kernel/paravirt.c             |  3 ---
 arch/arm64/include/asm/paravirt.h      |  4 ----
 arch/arm64/kernel/paravirt.c           |  4 +---
 arch/loongarch/include/asm/paravirt.h  |  3 ---
 arch/loongarch/kernel/paravirt.c       |  3 +--
 arch/powerpc/include/asm/paravirt.h    |  3 ---
 arch/powerpc/platforms/pseries/setup.c |  4 +---
 arch/riscv/include/asm/paravirt.h      |  4 ----
 arch/riscv/kernel/paravirt.c           |  4 +---
 arch/x86/include/asm/paravirt.h        |  4 ----
 arch/x86/kernel/cpu/vmware.c           |  1 +
 arch/x86/kernel/kvm.c                  |  1 +
 arch/x86/kernel/paravirt.c             |  3 ---
 drivers/xen/time.c                     |  1 +
 include/linux/sched/cputime.h          | 18 ++++++++++++++++++
 kernel/sched/core.c                    |  5 +++++
 kernel/sched/cputime.c                 | 13 +++++++++++++
 kernel/sched/sched.h                   |  2 +-
 20 files changed, 47 insertions(+), 40 deletions(-)

diff --git a/arch/Kconfig b/arch/Kconfig
index 61130b88964b..e3e39c2efa90 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -1059,6 +1059,9 @@ config HAVE_IRQ_TIME_ACCOUNTING
 	  Archs need to ensure they use a high enough resolution clock to
 	  support irq time accounting and then call enable_sched_clock_irqtime().
 
+config HAVE_PV_STEAL_CLOCK_GEN
+	bool
+
 config HAVE_MOVE_PUD
 	bool
 	help
diff --git a/arch/arm/include/asm/paravirt.h b/arch/arm/include/asm/paravirt.h
index 95d5b0d625cd..69da4bdcf856 100644
--- a/arch/arm/include/asm/paravirt.h
+++ b/arch/arm/include/asm/paravirt.h
@@ -5,10 +5,6 @@
 #ifdef CONFIG_PARAVIRT
 #include <linux/static_call_types.h>
 
-struct static_key;
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
-
 u64 dummy_steal_clock(int cpu);
 
 DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
diff --git a/arch/arm/kernel/paravirt.c b/arch/arm/kernel/paravirt.c
index 7dd9806369fb..3895a5578852 100644
--- a/arch/arm/kernel/paravirt.c
+++ b/arch/arm/kernel/paravirt.c
@@ -12,9 +12,6 @@
 #include <linux/static_call.h>
 #include <asm/paravirt.h>
 
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
-
 static u64 native_steal_clock(int cpu)
 {
 	return 0;
diff --git a/arch/arm64/include/asm/paravirt.h b/arch/arm64/include/asm/paravirt.h
index 9aa193e0e8f2..c9f7590baacb 100644
--- a/arch/arm64/include/asm/paravirt.h
+++ b/arch/arm64/include/asm/paravirt.h
@@ -5,10 +5,6 @@
 #ifdef CONFIG_PARAVIRT
 #include <linux/static_call_types.h>
 
-struct static_key;
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
-
 u64 dummy_steal_clock(int cpu);
 
 DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
diff --git a/arch/arm64/kernel/paravirt.c b/arch/arm64/kernel/paravirt.c
index aa718d6a9274..943b60ce12f4 100644
--- a/arch/arm64/kernel/paravirt.c
+++ b/arch/arm64/kernel/paravirt.c
@@ -19,14 +19,12 @@
 #include <linux/slab.h>
 #include <linux/types.h>
 #include <linux/static_call.h>
+#include <linux/sched/cputime.h>
 
 #include <asm/paravirt.h>
 #include <asm/pvclock-abi.h>
 #include <asm/smp_plat.h>
 
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
-
 static u64 native_steal_clock(int cpu)
 {
 	return 0;
diff --git a/arch/loongarch/include/asm/paravirt.h b/arch/loongarch/include/asm/paravirt.h
index 3f4323603e6a..d219ea0d98ac 100644
--- a/arch/loongarch/include/asm/paravirt.h
+++ b/arch/loongarch/include/asm/paravirt.h
@@ -5,9 +5,6 @@
 #ifdef CONFIG_PARAVIRT
 
 #include <linux/static_call_types.h>
-struct static_key;
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
 
 u64 dummy_steal_clock(int cpu);
 DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
diff --git a/arch/loongarch/kernel/paravirt.c b/arch/loongarch/kernel/paravirt.c
index b1b51f920b23..8caaa94fed1a 100644
--- a/arch/loongarch/kernel/paravirt.c
+++ b/arch/loongarch/kernel/paravirt.c
@@ -6,11 +6,10 @@
 #include <linux/kvm_para.h>
 #include <linux/reboot.h>
 #include <linux/static_call.h>
+#include <linux/sched/cputime.h>
 #include <asm/paravirt.h>
 
 static int has_steal_clock;
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
 static DEFINE_PER_CPU(struct kvm_steal_time, steal_time) __aligned(64);
 DEFINE_STATIC_KEY_FALSE(virt_spin_lock_key);
 
diff --git a/arch/powerpc/include/asm/paravirt.h b/arch/powerpc/include/asm/paravirt.h
index b78b82d66057..92343a23ad15 100644
--- a/arch/powerpc/include/asm/paravirt.h
+++ b/arch/powerpc/include/asm/paravirt.h
@@ -23,9 +23,6 @@ static inline bool is_shared_processor(void)
 }
 
 #ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
-
 u64 pseries_paravirt_steal_clock(int cpu);
 
 static inline u64 paravirt_steal_clock(int cpu)
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index b10a25325238..50b26ed8432d 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -42,6 +42,7 @@
 #include <linux/memblock.h>
 #include <linux/swiotlb.h>
 #include <linux/seq_buf.h>
+#include <linux/sched/cputime.h>
 
 #include <asm/mmu.h>
 #include <asm/processor.h>
@@ -83,9 +84,6 @@ DEFINE_STATIC_KEY_FALSE(shared_processor);
 EXPORT_SYMBOL(shared_processor);
 
 #ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
-
 static bool steal_acc = true;
 static int __init parse_no_stealacc(char *arg)
 {
diff --git a/arch/riscv/include/asm/paravirt.h b/arch/riscv/include/asm/paravirt.h
index c0abde70fc2c..17e5e39c72c0 100644
--- a/arch/riscv/include/asm/paravirt.h
+++ b/arch/riscv/include/asm/paravirt.h
@@ -5,10 +5,6 @@
 #ifdef CONFIG_PARAVIRT
 #include <linux/static_call_types.h>
 
-struct static_key;
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
-
 u64 dummy_steal_clock(int cpu);
 
 DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
diff --git a/arch/riscv/kernel/paravirt.c b/arch/riscv/kernel/paravirt.c
index fa6b0339a65d..d3c334f16172 100644
--- a/arch/riscv/kernel/paravirt.c
+++ b/arch/riscv/kernel/paravirt.c
@@ -16,15 +16,13 @@
 #include <linux/printk.h>
 #include <linux/static_call.h>
 #include <linux/types.h>
+#include <linux/sched/cputime.h>
 
 #include <asm/barrier.h>
 #include <asm/page.h>
 #include <asm/paravirt.h>
 #include <asm/sbi.h>
 
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
-
 static u64 native_steal_clock(int cpu)
 {
 	return 0;
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index 1344d2fb2b86..0ef797ea8440 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -30,10 +30,6 @@ static __always_inline u64 paravirt_sched_clock(void)
 	return static_call(pv_sched_clock)();
 }
 
-struct static_key;
-extern struct static_key paravirt_steal_enabled;
-extern struct static_key paravirt_steal_rq_enabled;
-
 __visible void __native_queued_spin_unlock(struct qspinlock *lock);
 bool pv_is_native_spin_unlock(void);
 __visible bool __native_vcpu_is_preempted(long cpu);
diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c
index cb3f900c46fc..a3e6936839b1 100644
--- a/arch/x86/kernel/cpu/vmware.c
+++ b/arch/x86/kernel/cpu/vmware.c
@@ -29,6 +29,7 @@
 #include <linux/efi.h>
 #include <linux/reboot.h>
 #include <linux/static_call.h>
+#include <linux/sched/cputime.h>
 #include <asm/div64.h>
 #include <asm/x86_init.h>
 #include <asm/hypervisor.h>
diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index b67d7c59dca0..d54fd2bc0402 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -29,6 +29,7 @@
 #include <linux/syscore_ops.h>
 #include <linux/cc_platform.h>
 #include <linux/efi.h>
+#include <linux/sched/cputime.h>
 #include <asm/timer.h>
 #include <asm/cpu.h>
 #include <asm/traps.h>
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index ab3e172dcc69..a3ba4747be1c 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -60,9 +60,6 @@ void __init native_pv_lock_init(void)
 		static_branch_enable(&virt_spin_lock_key);
 }
 
-struct static_key paravirt_steal_enabled;
-struct static_key paravirt_steal_rq_enabled;
-
 static u64 native_steal_clock(int cpu)
 {
 	return 0;
diff --git a/drivers/xen/time.c b/drivers/xen/time.c
index 5683383d2305..d360ded2ef39 100644
--- a/drivers/xen/time.c
+++ b/drivers/xen/time.c
@@ -8,6 +8,7 @@
 #include <linux/gfp.h>
 #include <linux/slab.h>
 #include <linux/static_call.h>
+#include <linux/sched/cputime.h>
 
 #include <asm/paravirt.h>
 #include <asm/xen/hypervisor.h>
diff --git a/include/linux/sched/cputime.h b/include/linux/sched/cputime.h
index 5f8fd5b24a2e..e90efaf6d26e 100644
--- a/include/linux/sched/cputime.h
+++ b/include/linux/sched/cputime.h
@@ -2,6 +2,7 @@
 #ifndef _LINUX_SCHED_CPUTIME_H
 #define _LINUX_SCHED_CPUTIME_H
 
+#include <linux/static_call_types.h>
 #include <linux/sched/signal.h>
 
 /*
@@ -180,4 +181,21 @@ static inline void prev_cputime_init(struct prev_cputime *prev)
 extern unsigned long long
 task_sched_runtime(struct task_struct *task);
 
+#ifdef CONFIG_PARAVIRT
+struct static_key;
+extern struct static_key paravirt_steal_enabled;
+extern struct static_key paravirt_steal_rq_enabled;
+
+#ifdef CONFIG_HAVE_PV_STEAL_CLOCK_GEN
+u64 dummy_steal_clock(int cpu);
+
+DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
+
+static inline u64 paravirt_steal_clock(int cpu)
+{
+	return static_call(pv_steal_clock)(cpu);
+}
+#endif
+#endif
+
 #endif /* _LINUX_SCHED_CPUTIME_H */
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index f754a60de848..9e7472646233 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -769,6 +769,11 @@ struct rq *task_rq_lock(struct task_struct *p, struct rq_flags *rf)
  * RQ-clock updating methods:
  */
 
+/* Use CONFIG_PARAVIRT as this will avoid more #ifdef in arch code. */
+#ifdef CONFIG_PARAVIRT
+struct static_key paravirt_steal_rq_enabled;
+#endif
+
 static void update_rq_clock_task(struct rq *rq, s64 delta)
 {
 /*
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 7097de2c8cda..ed8f71e08047 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -251,6 +251,19 @@ void __account_forceidle_time(struct task_struct *p, u64 delta)
  * ticks are not redelivered later. Due to that, this function may on
  * occasion account more time than the calling functions think elapsed.
  */
+#ifdef CONFIG_PARAVIRT
+struct static_key paravirt_steal_enabled;
+
+#ifdef CONFIG_HAVE_PV_STEAL_CLOCK_GEN
+static u64 native_steal_clock(int cpu)
+{
+	return 0;
+}
+
+DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
+#endif
+#endif
+
 static __always_inline u64 steal_account_process_time(u64 maxtime)
 {
 #ifdef CONFIG_PARAVIRT
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index ccd0d92ff7d1..bd1e5b55ba32 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -82,7 +82,7 @@ struct rt_rq;
 struct sched_group;
 struct cpuidle_state;
 
-#ifdef CONFIG_PARAVIRT
+#if defined(CONFIG_PARAVIRT) && !defined(CONFIG_HAVE_PV_STEAL_CLOCK_GEN)
 # include <asm/paravirt.h>
 #endif
 
-- 
2.51.0

[PATCH v4 07/21] arm/paravirt: Use common code for paravirt_steal_clock()

[Juergen Gross]

Remove the arch specific variant of paravirt_steal_clock() and use
the common one instead.

This allows to remove paravirt.c and paravirt.h from arch/arm.

Until all archs supporting Xen have been switched to the common code
of paravirt_steal_clock(), drivers/xen/time.c needs to include
asm/paravirt.h for those archs, while this is not necessary for arm
any longer.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/arm/Kconfig                |  1 +
 arch/arm/include/asm/paravirt.h | 18 ------------------
 arch/arm/kernel/Makefile        |  1 -
 arch/arm/kernel/paravirt.c      | 20 --------------------
 drivers/xen/time.c              |  2 ++
 5 files changed, 3 insertions(+), 39 deletions(-)
 delete mode 100644 arch/arm/include/asm/paravirt.h
 delete mode 100644 arch/arm/kernel/paravirt.c

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 4fb985b76e97..4a3044acebd6 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -1322,6 +1322,7 @@ config UACCESS_WITH_MEMCPY
 
 config PARAVIRT
 	bool "Enable paravirtualization code"
+	select HAVE_PV_STEAL_CLOCK_GEN
 	help
 	  This changes the kernel so it can modify itself when it is run
 	  under a hypervisor, potentially improving performance significantly
diff --git a/arch/arm/include/asm/paravirt.h b/arch/arm/include/asm/paravirt.h
deleted file mode 100644
index 69da4bdcf856..000000000000
--- a/arch/arm/include/asm/paravirt.h
+++ /dev/null
@@ -1,18 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _ASM_ARM_PARAVIRT_H
-#define _ASM_ARM_PARAVIRT_H
-
-#ifdef CONFIG_PARAVIRT
-#include <linux/static_call_types.h>
-
-u64 dummy_steal_clock(int cpu);
-
-DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
-
-static inline u64 paravirt_steal_clock(int cpu)
-{
-	return static_call(pv_steal_clock)(cpu);
-}
-#endif
-
-#endif
diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile
index afc9de7ef9a1..b36cf0cfd4a7 100644
--- a/arch/arm/kernel/Makefile
+++ b/arch/arm/kernel/Makefile
@@ -83,7 +83,6 @@ AFLAGS_iwmmxt.o			:= -Wa,-mcpu=iwmmxt
 obj-$(CONFIG_ARM_CPU_TOPOLOGY)  += topology.o
 obj-$(CONFIG_VDSO)		+= vdso.o
 obj-$(CONFIG_EFI)		+= efi.o
-obj-$(CONFIG_PARAVIRT)	+= paravirt.o
 
 obj-y			+= head$(MMUEXT).o
 obj-$(CONFIG_DEBUG_LL)	+= debug.o
diff --git a/arch/arm/kernel/paravirt.c b/arch/arm/kernel/paravirt.c
deleted file mode 100644
index 3895a5578852..000000000000
--- a/arch/arm/kernel/paravirt.c
+++ /dev/null
@@ -1,20 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/*
- *
- * Copyright (C) 2013 Citrix Systems
- *
- * Author: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
- */
-
-#include <linux/export.h>
-#include <linux/jump_label.h>
-#include <linux/types.h>
-#include <linux/static_call.h>
-#include <asm/paravirt.h>
-
-static u64 native_steal_clock(int cpu)
-{
-	return 0;
-}
-
-DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
diff --git a/drivers/xen/time.c b/drivers/xen/time.c
index d360ded2ef39..53b12f5ac465 100644
--- a/drivers/xen/time.c
+++ b/drivers/xen/time.c
@@ -10,7 +10,9 @@
 #include <linux/static_call.h>
 #include <linux/sched/cputime.h>
 
+#ifndef CONFIG_HAVE_PV_STEAL_CLOCK_GEN
 #include <asm/paravirt.h>
+#endif
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/hypercall.h>
 
-- 
2.51.0

[PATCH v4 08/21] arm64/paravirt: Use common code for paravirt_steal_clock()

[Juergen Gross]

Remove the arch specific variant of paravirt_steal_clock() and use
the common one instead.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/arm64/Kconfig                |  1 +
 arch/arm64/include/asm/paravirt.h | 10 ----------
 arch/arm64/kernel/paravirt.c      |  7 -------
 3 files changed, 1 insertion(+), 17 deletions(-)

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 6663ffd23f25..3a463027538e 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -1560,6 +1560,7 @@ config CC_HAVE_SHADOW_CALL_STACK
 
 config PARAVIRT
 	bool "Enable paravirtualization code"
+	select HAVE_PV_STEAL_CLOCK_GEN
 	help
 	  This changes the kernel so it can modify itself when it is run
 	  under a hypervisor, potentially improving performance significantly
diff --git a/arch/arm64/include/asm/paravirt.h b/arch/arm64/include/asm/paravirt.h
index c9f7590baacb..cb037e742372 100644
--- a/arch/arm64/include/asm/paravirt.h
+++ b/arch/arm64/include/asm/paravirt.h
@@ -3,16 +3,6 @@
 #define _ASM_ARM64_PARAVIRT_H
 
 #ifdef CONFIG_PARAVIRT
-#include <linux/static_call_types.h>
-
-u64 dummy_steal_clock(int cpu);
-
-DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
-
-static inline u64 paravirt_steal_clock(int cpu)
-{
-	return static_call(pv_steal_clock)(cpu);
-}
 
 int __init pv_time_init(void);
 
diff --git a/arch/arm64/kernel/paravirt.c b/arch/arm64/kernel/paravirt.c
index 943b60ce12f4..572efb96b23f 100644
--- a/arch/arm64/kernel/paravirt.c
+++ b/arch/arm64/kernel/paravirt.c
@@ -25,13 +25,6 @@
 #include <asm/pvclock-abi.h>
 #include <asm/smp_plat.h>
 
-static u64 native_steal_clock(int cpu)
-{
-	return 0;
-}
-
-DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
-
 struct pv_time_stolen_time_region {
 	struct pvclock_vcpu_stolen_time __rcu *kaddr;
 };
-- 
2.51.0

[PATCH v4 09/21] loongarch/paravirt: Use common code for paravirt_steal_clock()

[Juergen Gross]

Remove the arch specific variant of paravirt_steal_clock() and use
the common one instead.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
---
 arch/loongarch/Kconfig                |  1 +
 arch/loongarch/include/asm/paravirt.h | 10 ----------
 arch/loongarch/kernel/paravirt.c      |  7 -------
 3 files changed, 1 insertion(+), 17 deletions(-)

diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig
index 5b1116733d88..030fd487567c 100644
--- a/arch/loongarch/Kconfig
+++ b/arch/loongarch/Kconfig
@@ -682,6 +682,7 @@ source "kernel/livepatch/Kconfig"
 config PARAVIRT
 	bool "Enable paravirtualization code"
 	depends on AS_HAS_LVZ_EXTENSION
+	select HAVE_PV_STEAL_CLOCK_GEN
 	help
 	  This changes the kernel so it can modify itself when it is run
 	  under a hypervisor, potentially improving performance significantly
diff --git a/arch/loongarch/include/asm/paravirt.h b/arch/loongarch/include/asm/paravirt.h
index d219ea0d98ac..0111f0ad5f73 100644
--- a/arch/loongarch/include/asm/paravirt.h
+++ b/arch/loongarch/include/asm/paravirt.h
@@ -4,16 +4,6 @@
 
 #ifdef CONFIG_PARAVIRT
 
-#include <linux/static_call_types.h>
-
-u64 dummy_steal_clock(int cpu);
-DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
-
-static inline u64 paravirt_steal_clock(int cpu)
-{
-	return static_call(pv_steal_clock)(cpu);
-}
-
 int __init pv_ipi_init(void);
 int __init pv_time_init(void);
 int __init pv_spinlock_init(void);
diff --git a/arch/loongarch/kernel/paravirt.c b/arch/loongarch/kernel/paravirt.c
index 8caaa94fed1a..c5e526098c0b 100644
--- a/arch/loongarch/kernel/paravirt.c
+++ b/arch/loongarch/kernel/paravirt.c
@@ -13,13 +13,6 @@ static int has_steal_clock;
 static DEFINE_PER_CPU(struct kvm_steal_time, steal_time) __aligned(64);
 DEFINE_STATIC_KEY_FALSE(virt_spin_lock_key);
 
-static u64 native_steal_clock(int cpu)
-{
-	return 0;
-}
-
-DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
-
 static bool steal_acc = true;
 
 static int __init parse_no_stealacc(char *arg)
-- 
2.51.0

[PATCH v4 10/21] riscv/paravirt: Use common code for paravirt_steal_clock()

[Juergen Gross]

Remove the arch specific variant of paravirt_steal_clock() and use
the common one instead.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
---
 arch/riscv/Kconfig                |  1 +
 arch/riscv/include/asm/paravirt.h | 10 ----------
 arch/riscv/kernel/paravirt.c      |  7 -------
 3 files changed, 1 insertion(+), 17 deletions(-)

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index fadec20b87a8..6455c57e03fb 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -1095,6 +1095,7 @@ config COMPAT
 config PARAVIRT
 	bool "Enable paravirtualization code"
 	depends on RISCV_SBI
+	select HAVE_PV_STEAL_CLOCK_GEN
 	help
 	  This changes the kernel so it can modify itself when it is run
 	  under a hypervisor, potentially improving performance significantly
diff --git a/arch/riscv/include/asm/paravirt.h b/arch/riscv/include/asm/paravirt.h
index 17e5e39c72c0..c49c55b266f3 100644
--- a/arch/riscv/include/asm/paravirt.h
+++ b/arch/riscv/include/asm/paravirt.h
@@ -3,16 +3,6 @@
 #define _ASM_RISCV_PARAVIRT_H
 
 #ifdef CONFIG_PARAVIRT
-#include <linux/static_call_types.h>
-
-u64 dummy_steal_clock(int cpu);
-
-DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
-
-static inline u64 paravirt_steal_clock(int cpu)
-{
-	return static_call(pv_steal_clock)(cpu);
-}
 
 int __init pv_time_init(void);
 
diff --git a/arch/riscv/kernel/paravirt.c b/arch/riscv/kernel/paravirt.c
index d3c334f16172..5f56be79cd06 100644
--- a/arch/riscv/kernel/paravirt.c
+++ b/arch/riscv/kernel/paravirt.c
@@ -23,13 +23,6 @@
 #include <asm/paravirt.h>
 #include <asm/sbi.h>
 
-static u64 native_steal_clock(int cpu)
-{
-	return 0;
-}
-
-DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
-
 static bool steal_acc = true;
 static int __init parse_no_stealacc(char *arg)
 {
-- 
2.51.0

[PATCH v4 11/21] x86/paravirt: Use common code for paravirt_steal_clock()

[Juergen Gross]

Remove the arch specific variant of paravirt_steal_clock() and use
the common one instead.

With all archs supporting Xen now having been switched to the common
variant, including paravirt.h can be dropped from drivers/xen/time.c.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/x86/Kconfig                | 1 +
 arch/x86/include/asm/paravirt.h | 7 -------
 arch/x86/kernel/paravirt.c      | 6 ------
 arch/x86/xen/time.c             | 1 +
 drivers/xen/time.c              | 3 ---
 5 files changed, 2 insertions(+), 16 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index aa279fec5c1f..eee6541a5765 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -795,6 +795,7 @@ if HYPERVISOR_GUEST
 config PARAVIRT
 	bool "Enable paravirtualization code"
 	depends on HAVE_STATIC_CALL
+	select HAVE_PV_STEAL_CLOCK_GEN
 	help
 	  This changes the kernel so it can modify itself when it is run
 	  under a hypervisor, potentially improving performance significantly
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index 0ef797ea8440..766a7cee3d64 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -17,10 +17,8 @@
 #include <linux/static_call_types.h>
 #include <asm/frame.h>
 
-u64 dummy_steal_clock(int cpu);
 u64 dummy_sched_clock(void);
 
-DECLARE_STATIC_CALL(pv_steal_clock, dummy_steal_clock);
 DECLARE_STATIC_CALL(pv_sched_clock, dummy_sched_clock);
 
 void paravirt_set_sched_clock(u64 (*func)(void));
@@ -35,11 +33,6 @@ bool pv_is_native_spin_unlock(void);
 __visible bool __native_vcpu_is_preempted(long cpu);
 bool pv_is_native_vcpu_is_preempted(void);
 
-static inline u64 paravirt_steal_clock(int cpu)
-{
-	return static_call(pv_steal_clock)(cpu);
-}
-
 #ifdef CONFIG_PARAVIRT_SPINLOCKS
 void __init paravirt_set_cap(void);
 #endif
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index a3ba4747be1c..42991d471bf3 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -60,12 +60,6 @@ void __init native_pv_lock_init(void)
 		static_branch_enable(&virt_spin_lock_key);
 }
 
-static u64 native_steal_clock(int cpu)
-{
-	return 0;
-}
-
-DEFINE_STATIC_CALL(pv_steal_clock, native_steal_clock);
 DEFINE_STATIC_CALL(pv_sched_clock, native_sched_clock);
 
 void paravirt_set_sched_clock(u64 (*func)(void))
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index 96521b1874ac..e4754b2fa900 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -16,6 +16,7 @@
 #include <linux/slab.h>
 #include <linux/pvclock_gtod.h>
 #include <linux/timekeeper_internal.h>
+#include <linux/sched/cputime.h>
 
 #include <asm/pvclock.h>
 #include <asm/xen/hypervisor.h>
diff --git a/drivers/xen/time.c b/drivers/xen/time.c
index 53b12f5ac465..0b18d8a5a2dd 100644
--- a/drivers/xen/time.c
+++ b/drivers/xen/time.c
@@ -10,9 +10,6 @@
 #include <linux/static_call.h>
 #include <linux/sched/cputime.h>
 
-#ifndef CONFIG_HAVE_PV_STEAL_CLOCK_GEN
-#include <asm/paravirt.h>
-#endif
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/hypercall.h>
 
-- 
2.51.0

[PATCH v4 12/21] x86/paravirt: Move paravirt_sched_clock() related code into tsc.c

[Juergen Gross]

The only user of paravirt_sched_clock() is in tsc.c, so move the code
from paravirt.c and paravirt.h to tsc.c.

Signed-off-by: Juergen Gross <jgross@suse.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 arch/x86/include/asm/paravirt.h    | 12 ------------
 arch/x86/include/asm/timer.h       |  1 +
 arch/x86/kernel/kvmclock.c         |  1 +
 arch/x86/kernel/paravirt.c         |  7 -------
 arch/x86/kernel/tsc.c              | 10 +++++++++-
 arch/x86/xen/time.c                |  1 +
 drivers/clocksource/hyperv_timer.c |  2 ++
 7 files changed, 14 insertions(+), 20 deletions(-)

diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index 766a7cee3d64..b69e75a5c872 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -14,20 +14,8 @@
 #ifndef __ASSEMBLER__
 #include <linux/types.h>
 #include <linux/cpumask.h>
-#include <linux/static_call_types.h>
 #include <asm/frame.h>
 
-u64 dummy_sched_clock(void);
-
-DECLARE_STATIC_CALL(pv_sched_clock, dummy_sched_clock);
-
-void paravirt_set_sched_clock(u64 (*func)(void));
-
-static __always_inline u64 paravirt_sched_clock(void)
-{
-	return static_call(pv_sched_clock)();
-}
-
 __visible void __native_queued_spin_unlock(struct qspinlock *lock);
 bool pv_is_native_spin_unlock(void);
 __visible bool __native_vcpu_is_preempted(long cpu);
diff --git a/arch/x86/include/asm/timer.h b/arch/x86/include/asm/timer.h
index 23baf8c9b34c..fda18bcb19b4 100644
--- a/arch/x86/include/asm/timer.h
+++ b/arch/x86/include/asm/timer.h
@@ -12,6 +12,7 @@ extern void recalibrate_cpu_khz(void);
 extern int no_timer_check;
 
 extern bool using_native_sched_clock(void);
+void paravirt_set_sched_clock(u64 (*func)(void));
 
 /*
  * We use the full linear equation: f(x) = a + b*x, in order to allow
diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c
index ca0a49eeac4a..b5991d53fc0e 100644
--- a/arch/x86/kernel/kvmclock.c
+++ b/arch/x86/kernel/kvmclock.c
@@ -19,6 +19,7 @@
 #include <linux/cc_platform.h>
 
 #include <asm/hypervisor.h>
+#include <asm/timer.h>
 #include <asm/x86_init.h>
 #include <asm/kvmclock.h>
 
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index 42991d471bf3..4e37db8073f9 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -60,13 +60,6 @@ void __init native_pv_lock_init(void)
 		static_branch_enable(&virt_spin_lock_key);
 }
 
-DEFINE_STATIC_CALL(pv_sched_clock, native_sched_clock);
-
-void paravirt_set_sched_clock(u64 (*func)(void))
-{
-	static_call_update(pv_sched_clock, func);
-}
-
 static noinstr void pv_native_safe_halt(void)
 {
 	native_safe_halt();
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 87e749106dda..554b54783a04 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -266,19 +266,27 @@ u64 native_sched_clock_from_tsc(u64 tsc)
 /* We need to define a real function for sched_clock, to override the
    weak default version */
 #ifdef CONFIG_PARAVIRT
+DEFINE_STATIC_CALL(pv_sched_clock, native_sched_clock);
+
 noinstr u64 sched_clock_noinstr(void)
 {
-	return paravirt_sched_clock();
+	return static_call(pv_sched_clock)();
 }
 
 bool using_native_sched_clock(void)
 {
 	return static_call_query(pv_sched_clock) == native_sched_clock;
 }
+
+void paravirt_set_sched_clock(u64 (*func)(void))
+{
+	static_call_update(pv_sched_clock, func);
+}
 #else
 u64 sched_clock_noinstr(void) __attribute__((alias("native_sched_clock")));
 
 bool using_native_sched_clock(void) { return true; }
+void paravirt_set_sched_clock(u64 (*func)(void)) { }
 #endif
 
 notrace u64 sched_clock(void)
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index e4754b2fa900..6f9f665bb7ae 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -19,6 +19,7 @@
 #include <linux/sched/cputime.h>
 
 #include <asm/pvclock.h>
+#include <asm/timer.h>
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/hypercall.h>
 #include <asm/xen/cpuid.h>
diff --git a/drivers/clocksource/hyperv_timer.c b/drivers/clocksource/hyperv_timer.c
index 10356d4ec55c..e9f5034a1bc8 100644
--- a/drivers/clocksource/hyperv_timer.c
+++ b/drivers/clocksource/hyperv_timer.c
@@ -535,6 +535,8 @@ static __always_inline void hv_setup_sched_clock(void *sched_clock)
 	sched_clock_register(sched_clock, 64, NSEC_PER_SEC);
 }
 #elif defined CONFIG_PARAVIRT
+#include <asm/timer.h>
+
 static __always_inline void hv_setup_sched_clock(void *sched_clock)
 {
 	/* We're on x86/x64 *and* using PV ops */
-- 
2.51.0

[PATCH v4 13/21] x86/paravirt: Introduce new paravirt-base.h header

[Juergen Gross]

Move the pv_info related definitions and the declarations of the global
paravirt function primitives into a new header file paravirt-base.h.

This enables to use that header instead of paravirt_types.h in
ptrace.h.

Additionally it is in preparation of reducing include hell with
paravirt enabled.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
V2:
- new patch
---
 arch/x86/include/asm/paravirt-base.h  | 29 +++++++++++++++++++++++++++
 arch/x86/include/asm/paravirt.h       |  4 +++-
 arch/x86/include/asm/paravirt_types.h | 23 +--------------------
 arch/x86/include/asm/ptrace.h         |  2 +-
 4 files changed, 34 insertions(+), 24 deletions(-)
 create mode 100644 arch/x86/include/asm/paravirt-base.h

diff --git a/arch/x86/include/asm/paravirt-base.h b/arch/x86/include/asm/paravirt-base.h
new file mode 100644
index 000000000000..3827ea20de18
--- /dev/null
+++ b/arch/x86/include/asm/paravirt-base.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef _ASM_X86_PARAVIRT_BASE_H
+#define _ASM_X86_PARAVIRT_BASE_H
+
+/*
+ * Wrapper type for pointers to code which uses the non-standard
+ * calling convention.  See PV_CALL_SAVE_REGS_THUNK below.
+ */
+struct paravirt_callee_save {
+	void *func;
+};
+
+struct pv_info {
+#ifdef CONFIG_PARAVIRT_XXL
+	u16 extra_user_64bit_cs;  /* __USER_CS if none */
+#endif
+	const char *name;
+};
+
+void default_banner(void);
+extern struct pv_info pv_info;
+unsigned long paravirt_ret0(void);
+#ifdef CONFIG_PARAVIRT_XXL
+u64 _paravirt_ident_64(u64);
+#endif
+#define paravirt_nop	((void *)nop_func)
+
+#endif /* _ASM_X86_PARAVIRT_BASE_H */
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index b69e75a5c872..62399f5d037d 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -4,6 +4,9 @@
 /* Various instructions on x86 need to be replaced for
  * para-virtualization: those hooks are defined here. */
 
+#ifndef __ASSEMBLER__
+#include <asm/paravirt-base.h>
+#endif
 #include <asm/paravirt_types.h>
 
 #ifdef CONFIG_PARAVIRT
@@ -601,7 +604,6 @@ static __always_inline unsigned long arch_local_irq_save(void)
 #undef PVOP_VCALL4
 #undef PVOP_CALL4
 
-extern void default_banner(void);
 void native_pv_lock_init(void) __init;
 
 #else  /* __ASSEMBLER__ */
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 7acff40cc159..148d157e2a4a 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -7,6 +7,7 @@
 #ifndef __ASSEMBLER__
 #include <linux/types.h>
 
+#include <asm/paravirt-base.h>
 #include <asm/desc_defs.h>
 #include <asm/pgtable_types.h>
 #include <asm/nospec-branch.h>
@@ -18,23 +19,6 @@ struct cpumask;
 struct flush_tlb_info;
 struct vm_area_struct;
 
-/*
- * Wrapper type for pointers to code which uses the non-standard
- * calling convention.  See PV_CALL_SAVE_REGS_THUNK below.
- */
-struct paravirt_callee_save {
-	void *func;
-};
-
-/* general info */
-struct pv_info {
-#ifdef CONFIG_PARAVIRT_XXL
-	u16 extra_user_64bit_cs;  /* __USER_CS if none */
-#endif
-
-	const char *name;
-};
-
 #ifdef CONFIG_PARAVIRT_XXL
 struct pv_lazy_ops {
 	/* Set deferred update mode, used for batching operations. */
@@ -226,7 +210,6 @@ struct paravirt_patch_template {
 	struct pv_lock_ops	lock;
 } __no_randomize_layout;
 
-extern struct pv_info pv_info;
 extern struct paravirt_patch_template pv_ops;
 
 #define paravirt_ptr(op)	[paravirt_opptr] "m" (pv_ops.op)
@@ -497,17 +480,13 @@ extern struct paravirt_patch_template pv_ops;
 	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2),	\
 		     PVOP_CALL_ARG3(arg3), PVOP_CALL_ARG4(arg4))
 
-unsigned long paravirt_ret0(void);
 #ifdef CONFIG_PARAVIRT_XXL
-u64 _paravirt_ident_64(u64);
 unsigned long pv_native_save_fl(void);
 void pv_native_irq_disable(void);
 void pv_native_irq_enable(void);
 unsigned long pv_native_read_cr2(void);
 #endif
 
-#define paravirt_nop	((void *)nop_func)
-
 #endif	/* __ASSEMBLER__ */
 
 #define ALT_NOT_XEN	ALT_NOT(X86_FEATURE_XENPV)
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 50f75467f73d..fe2dab7d74e3 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -172,7 +172,7 @@ struct pt_regs {
 #endif /* !__i386__ */
 
 #ifdef CONFIG_PARAVIRT
-#include <asm/paravirt_types.h>
+#include <asm/paravirt-base.h>
 #endif
 
 #include <asm/proto.h>
-- 
2.51.0

[PATCH v4 14/21] x86/paravirt: Move pv_native_*() prototypes to paravirt.c

[Juergen Gross]

The only reason the pv_native_*() prototypes are needed is the complete
definition of those functions via an asm() statement, which makes it
impossible to have those functions as static ones.

Move the prototypes from paravirt_types.h into paravirt.c, which is the
only source referencing the functions.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
V2:
- new patch
---
 arch/x86/include/asm/paravirt_types.h | 7 -------
 arch/x86/kernel/paravirt.c            | 5 +++++
 2 files changed, 5 insertions(+), 7 deletions(-)

diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 148d157e2a4a..1e50f13e6543 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -480,13 +480,6 @@ extern struct paravirt_patch_template pv_ops;
 	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2),	\
 		     PVOP_CALL_ARG3(arg3), PVOP_CALL_ARG4(arg4))
 
-#ifdef CONFIG_PARAVIRT_XXL
-unsigned long pv_native_save_fl(void);
-void pv_native_irq_disable(void);
-void pv_native_irq_enable(void);
-unsigned long pv_native_read_cr2(void);
-#endif
-
 #endif	/* __ASSEMBLER__ */
 
 #define ALT_NOT_XEN	ALT_NOT(X86_FEATURE_XENPV)
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index 4e37db8073f9..5dfbd3f55792 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -45,6 +45,11 @@ void __init default_banner(void)
 }
 
 #ifdef CONFIG_PARAVIRT_XXL
+unsigned long pv_native_save_fl(void);
+void pv_native_irq_disable(void);
+void pv_native_irq_enable(void);
+unsigned long pv_native_read_cr2(void);
+
 DEFINE_ASM_FUNC(_paravirt_ident_64, "mov %rdi, %rax", .text);
 DEFINE_ASM_FUNC(pv_native_save_fl, "pushf; pop %rax", .noinstr.text);
 DEFINE_ASM_FUNC(pv_native_irq_disable, "cli", .noinstr.text);
-- 
2.51.0

[PATCH v4 19/21] x86/paravirt: Allow pv-calls outside paravirt.h

[Juergen Gross]

In order to prepare for defining paravirt functions outside of
paravirt.h, don't #undef the paravirt call macros.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
 arch/x86/include/asm/paravirt.h | 16 ----------------
 1 file changed, 16 deletions(-)

diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index 62399f5d037d..ba6b14b6f36a 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -588,22 +588,6 @@ static __always_inline unsigned long arch_local_irq_save(void)
 }
 #endif
 
-
-/* Make sure as little as possible of this mess escapes. */
-#undef PARAVIRT_CALL
-#undef __PVOP_CALL
-#undef __PVOP_VCALL
-#undef PVOP_VCALL0
-#undef PVOP_CALL0
-#undef PVOP_VCALL1
-#undef PVOP_CALL1
-#undef PVOP_VCALL2
-#undef PVOP_CALL2
-#undef PVOP_VCALL3
-#undef PVOP_CALL3
-#undef PVOP_VCALL4
-#undef PVOP_CALL4
-
 void native_pv_lock_init(void) __init;
 
 #else  /* __ASSEMBLER__ */
-- 
2.51.0

[PATCH v4 20/21] x86/paravirt: Specify pv_ops array in paravirt macros

[Juergen Gross]

In order to prepare having multiple pv_ops arrays, specify the array
in the paravirt macros.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
V2:
- new patch
---
 arch/x86/include/asm/paravirt.h       | 166 +++++++++++++-------------
 arch/x86/include/asm/paravirt_types.h | 140 +++++++++++-----------
 2 files changed, 153 insertions(+), 153 deletions(-)

diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index ba6b14b6f36a..ec274d13bae0 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -31,11 +31,11 @@ void __init paravirt_set_cap(void);
 /* The paravirtualized I/O functions */
 static inline void slow_down_io(void)
 {
-	PVOP_VCALL0(cpu.io_delay);
+	PVOP_VCALL0(pv_ops, cpu.io_delay);
 #ifdef REALLY_SLOW_IO
-	PVOP_VCALL0(cpu.io_delay);
-	PVOP_VCALL0(cpu.io_delay);
-	PVOP_VCALL0(cpu.io_delay);
+	PVOP_VCALL0(pv_ops, cpu.io_delay);
+	PVOP_VCALL0(pv_ops, cpu.io_delay);
+	PVOP_VCALL0(pv_ops, cpu.io_delay);
 #endif
 }
 
@@ -47,57 +47,57 @@ void native_flush_tlb_multi(const struct cpumask *cpumask,
 
 static inline void __flush_tlb_local(void)
 {
-	PVOP_VCALL0(mmu.flush_tlb_user);
+	PVOP_VCALL0(pv_ops, mmu.flush_tlb_user);
 }
 
 static inline void __flush_tlb_global(void)
 {
-	PVOP_VCALL0(mmu.flush_tlb_kernel);
+	PVOP_VCALL0(pv_ops, mmu.flush_tlb_kernel);
 }
 
 static inline void __flush_tlb_one_user(unsigned long addr)
 {
-	PVOP_VCALL1(mmu.flush_tlb_one_user, addr);
+	PVOP_VCALL1(pv_ops, mmu.flush_tlb_one_user, addr);
 }
 
 static inline void __flush_tlb_multi(const struct cpumask *cpumask,
 				      const struct flush_tlb_info *info)
 {
-	PVOP_VCALL2(mmu.flush_tlb_multi, cpumask, info);
+	PVOP_VCALL2(pv_ops, mmu.flush_tlb_multi, cpumask, info);
 }
 
 static inline void paravirt_arch_exit_mmap(struct mm_struct *mm)
 {
-	PVOP_VCALL1(mmu.exit_mmap, mm);
+	PVOP_VCALL1(pv_ops, mmu.exit_mmap, mm);
 }
 
 static inline void notify_page_enc_status_changed(unsigned long pfn,
 						  int npages, bool enc)
 {
-	PVOP_VCALL3(mmu.notify_page_enc_status_changed, pfn, npages, enc);
+	PVOP_VCALL3(pv_ops, mmu.notify_page_enc_status_changed, pfn, npages, enc);
 }
 
 static __always_inline void arch_safe_halt(void)
 {
-	PVOP_VCALL0(irq.safe_halt);
+	PVOP_VCALL0(pv_ops, irq.safe_halt);
 }
 
 static inline void halt(void)
 {
-	PVOP_VCALL0(irq.halt);
+	PVOP_VCALL0(pv_ops, irq.halt);
 }
 
 #ifdef CONFIG_PARAVIRT_XXL
 static inline void load_sp0(unsigned long sp0)
 {
-	PVOP_VCALL1(cpu.load_sp0, sp0);
+	PVOP_VCALL1(pv_ops, cpu.load_sp0, sp0);
 }
 
 /* The paravirtualized CPUID instruction. */
 static inline void __cpuid(unsigned int *eax, unsigned int *ebx,
 			   unsigned int *ecx, unsigned int *edx)
 {
-	PVOP_VCALL4(cpu.cpuid, eax, ebx, ecx, edx);
+	PVOP_VCALL4(pv_ops, cpu.cpuid, eax, ebx, ecx, edx);
 }
 
 /*
@@ -105,69 +105,69 @@ static inline void __cpuid(unsigned int *eax, unsigned int *ebx,
  */
 static __always_inline unsigned long paravirt_get_debugreg(int reg)
 {
-	return PVOP_CALL1(unsigned long, cpu.get_debugreg, reg);
+	return PVOP_CALL1(unsigned long, pv_ops, cpu.get_debugreg, reg);
 }
 #define get_debugreg(var, reg) var = paravirt_get_debugreg(reg)
 static __always_inline void set_debugreg(unsigned long val, int reg)
 {
-	PVOP_VCALL2(cpu.set_debugreg, reg, val);
+	PVOP_VCALL2(pv_ops, cpu.set_debugreg, reg, val);
 }
 
 static inline unsigned long read_cr0(void)
 {
-	return PVOP_CALL0(unsigned long, cpu.read_cr0);
+	return PVOP_CALL0(unsigned long, pv_ops, cpu.read_cr0);
 }
 
 static inline void write_cr0(unsigned long x)
 {
-	PVOP_VCALL1(cpu.write_cr0, x);
+	PVOP_VCALL1(pv_ops, cpu.write_cr0, x);
 }
 
 static __always_inline unsigned long read_cr2(void)
 {
-	return PVOP_ALT_CALLEE0(unsigned long, mmu.read_cr2,
+	return PVOP_ALT_CALLEE0(unsigned long, pv_ops, mmu.read_cr2,
 				"mov %%cr2, %%rax;", ALT_NOT_XEN);
 }
 
 static __always_inline void write_cr2(unsigned long x)
 {
-	PVOP_VCALL1(mmu.write_cr2, x);
+	PVOP_VCALL1(pv_ops, mmu.write_cr2, x);
 }
 
 static inline unsigned long __read_cr3(void)
 {
-	return PVOP_ALT_CALL0(unsigned long, mmu.read_cr3,
+	return PVOP_ALT_CALL0(unsigned long, pv_ops, mmu.read_cr3,
 			      "mov %%cr3, %%rax;", ALT_NOT_XEN);
 }
 
 static inline void write_cr3(unsigned long x)
 {
-	PVOP_ALT_VCALL1(mmu.write_cr3, x, "mov %%rdi, %%cr3", ALT_NOT_XEN);
+	PVOP_ALT_VCALL1(pv_ops, mmu.write_cr3, x, "mov %%rdi, %%cr3", ALT_NOT_XEN);
 }
 
 static inline void __write_cr4(unsigned long x)
 {
-	PVOP_VCALL1(cpu.write_cr4, x);
+	PVOP_VCALL1(pv_ops, cpu.write_cr4, x);
 }
 
 static inline u64 paravirt_read_msr(u32 msr)
 {
-	return PVOP_CALL1(u64, cpu.read_msr, msr);
+	return PVOP_CALL1(u64, pv_ops, cpu.read_msr, msr);
 }
 
 static inline void paravirt_write_msr(u32 msr, u64 val)
 {
-	PVOP_VCALL2(cpu.write_msr, msr, val);
+	PVOP_VCALL2(pv_ops, cpu.write_msr, msr, val);
 }
 
 static inline int paravirt_read_msr_safe(u32 msr, u64 *val)
 {
-	return PVOP_CALL2(int, cpu.read_msr_safe, msr, val);
+	return PVOP_CALL2(int, pv_ops, cpu.read_msr_safe, msr, val);
 }
 
 static inline int paravirt_write_msr_safe(u32 msr, u64 val)
 {
-	return PVOP_CALL2(int, cpu.write_msr_safe, msr, val);
+	return PVOP_CALL2(int, pv_ops, cpu.write_msr_safe, msr, val);
 }
 
 #define rdmsr(msr, val1, val2)			\
@@ -214,154 +214,154 @@ static __always_inline int rdmsrq_safe(u32 msr, u64 *p)
 
 static __always_inline u64 rdpmc(int counter)
 {
-	return PVOP_CALL1(u64, cpu.read_pmc, counter);
+	return PVOP_CALL1(u64, pv_ops, cpu.read_pmc, counter);
 }
 
 static inline void paravirt_alloc_ldt(struct desc_struct *ldt, unsigned entries)
 {
-	PVOP_VCALL2(cpu.alloc_ldt, ldt, entries);
+	PVOP_VCALL2(pv_ops, cpu.alloc_ldt, ldt, entries);
 }
 
 static inline void paravirt_free_ldt(struct desc_struct *ldt, unsigned entries)
 {
-	PVOP_VCALL2(cpu.free_ldt, ldt, entries);
+	PVOP_VCALL2(pv_ops, cpu.free_ldt, ldt, entries);
 }
 
 static inline void load_TR_desc(void)
 {
-	PVOP_VCALL0(cpu.load_tr_desc);
+	PVOP_VCALL0(pv_ops, cpu.load_tr_desc);
 }
 static inline void load_gdt(const struct desc_ptr *dtr)
 {
-	PVOP_VCALL1(cpu.load_gdt, dtr);
+	PVOP_VCALL1(pv_ops, cpu.load_gdt, dtr);
 }
 static inline void load_idt(const struct desc_ptr *dtr)
 {
-	PVOP_VCALL1(cpu.load_idt, dtr);
+	PVOP_VCALL1(pv_ops, cpu.load_idt, dtr);
 }
 static inline void set_ldt(const void *addr, unsigned entries)
 {
-	PVOP_VCALL2(cpu.set_ldt, addr, entries);
+	PVOP_VCALL2(pv_ops, cpu.set_ldt, addr, entries);
 }
 static inline unsigned long paravirt_store_tr(void)
 {
-	return PVOP_CALL0(unsigned long, cpu.store_tr);
+	return PVOP_CALL0(unsigned long, pv_ops, cpu.store_tr);
 }
 
 #define store_tr(tr)	((tr) = paravirt_store_tr())
 static inline void load_TLS(struct thread_struct *t, unsigned cpu)
 {
-	PVOP_VCALL2(cpu.load_tls, t, cpu);
+	PVOP_VCALL2(pv_ops, cpu.load_tls, t, cpu);
 }
 
 static inline void load_gs_index(unsigned int gs)
 {
-	PVOP_VCALL1(cpu.load_gs_index, gs);
+	PVOP_VCALL1(pv_ops, cpu.load_gs_index, gs);
 }
 
 static inline void write_ldt_entry(struct desc_struct *dt, int entry,
 				   const void *desc)
 {
-	PVOP_VCALL3(cpu.write_ldt_entry, dt, entry, desc);
+	PVOP_VCALL3(pv_ops, cpu.write_ldt_entry, dt, entry, desc);
 }
 
 static inline void write_gdt_entry(struct desc_struct *dt, int entry,
 				   void *desc, int type)
 {
-	PVOP_VCALL4(cpu.write_gdt_entry, dt, entry, desc, type);
+	PVOP_VCALL4(pv_ops, cpu.write_gdt_entry, dt, entry, desc, type);
 }
 
 static inline void write_idt_entry(gate_desc *dt, int entry, const gate_desc *g)
 {
-	PVOP_VCALL3(cpu.write_idt_entry, dt, entry, g);
+	PVOP_VCALL3(pv_ops, cpu.write_idt_entry, dt, entry, g);
 }
 
 #ifdef CONFIG_X86_IOPL_IOPERM
 static inline void tss_invalidate_io_bitmap(void)
 {
-	PVOP_VCALL0(cpu.invalidate_io_bitmap);
+	PVOP_VCALL0(pv_ops, cpu.invalidate_io_bitmap);
 }
 
 static inline void tss_update_io_bitmap(void)
 {
-	PVOP_VCALL0(cpu.update_io_bitmap);
+	PVOP_VCALL0(pv_ops, cpu.update_io_bitmap);
 }
 #endif
 
 static inline void paravirt_enter_mmap(struct mm_struct *next)
 {
-	PVOP_VCALL1(mmu.enter_mmap, next);
+	PVOP_VCALL1(pv_ops, mmu.enter_mmap, next);
 }
 
 static inline int paravirt_pgd_alloc(struct mm_struct *mm)
 {
-	return PVOP_CALL1(int, mmu.pgd_alloc, mm);
+	return PVOP_CALL1(int, pv_ops, mmu.pgd_alloc, mm);
 }
 
 static inline void paravirt_pgd_free(struct mm_struct *mm, pgd_t *pgd)
 {
-	PVOP_VCALL2(mmu.pgd_free, mm, pgd);
+	PVOP_VCALL2(pv_ops, mmu.pgd_free, mm, pgd);
 }
 
 static inline void paravirt_alloc_pte(struct mm_struct *mm, unsigned long pfn)
 {
-	PVOP_VCALL2(mmu.alloc_pte, mm, pfn);
+	PVOP_VCALL2(pv_ops, mmu.alloc_pte, mm, pfn);
 }
 static inline void paravirt_release_pte(unsigned long pfn)
 {
-	PVOP_VCALL1(mmu.release_pte, pfn);
+	PVOP_VCALL1(pv_ops, mmu.release_pte, pfn);
 }
 
 static inline void paravirt_alloc_pmd(struct mm_struct *mm, unsigned long pfn)
 {
-	PVOP_VCALL2(mmu.alloc_pmd, mm, pfn);
+	PVOP_VCALL2(pv_ops, mmu.alloc_pmd, mm, pfn);
 }
 
 static inline void paravirt_release_pmd(unsigned long pfn)
 {
-	PVOP_VCALL1(mmu.release_pmd, pfn);
+	PVOP_VCALL1(pv_ops, mmu.release_pmd, pfn);
 }
 
 static inline void paravirt_alloc_pud(struct mm_struct *mm, unsigned long pfn)
 {
-	PVOP_VCALL2(mmu.alloc_pud, mm, pfn);
+	PVOP_VCALL2(pv_ops, mmu.alloc_pud, mm, pfn);
 }
 static inline void paravirt_release_pud(unsigned long pfn)
 {
-	PVOP_VCALL1(mmu.release_pud, pfn);
+	PVOP_VCALL1(pv_ops, mmu.release_pud, pfn);
 }
 
 static inline void paravirt_alloc_p4d(struct mm_struct *mm, unsigned long pfn)
 {
-	PVOP_VCALL2(mmu.alloc_p4d, mm, pfn);
+	PVOP_VCALL2(pv_ops, mmu.alloc_p4d, mm, pfn);
 }
 
 static inline void paravirt_release_p4d(unsigned long pfn)
 {
-	PVOP_VCALL1(mmu.release_p4d, pfn);
+	PVOP_VCALL1(pv_ops, mmu.release_p4d, pfn);
 }
 
 static inline pte_t __pte(pteval_t val)
 {
-	return (pte_t) { PVOP_ALT_CALLEE1(pteval_t, mmu.make_pte, val,
+	return (pte_t) { PVOP_ALT_CALLEE1(pteval_t, pv_ops, mmu.make_pte, val,
 					  "mov %%rdi, %%rax", ALT_NOT_XEN) };
 }
 
 static inline pteval_t pte_val(pte_t pte)
 {
-	return PVOP_ALT_CALLEE1(pteval_t, mmu.pte_val, pte.pte,
+	return PVOP_ALT_CALLEE1(pteval_t, pv_ops, mmu.pte_val, pte.pte,
 				"mov %%rdi, %%rax", ALT_NOT_XEN);
 }
 
 static inline pgd_t __pgd(pgdval_t val)
 {
-	return (pgd_t) { PVOP_ALT_CALLEE1(pgdval_t, mmu.make_pgd, val,
+	return (pgd_t) { PVOP_ALT_CALLEE1(pgdval_t, pv_ops, mmu.make_pgd, val,
 					  "mov %%rdi, %%rax", ALT_NOT_XEN) };
 }
 
 static inline pgdval_t pgd_val(pgd_t pgd)
 {
-	return PVOP_ALT_CALLEE1(pgdval_t, mmu.pgd_val, pgd.pgd,
+	return PVOP_ALT_CALLEE1(pgdval_t, pv_ops, mmu.pgd_val, pgd.pgd,
 				"mov %%rdi, %%rax", ALT_NOT_XEN);
 }
 
@@ -371,7 +371,7 @@ static inline pte_t ptep_modify_prot_start(struct vm_area_struct *vma, unsigned
 {
 	pteval_t ret;
 
-	ret = PVOP_CALL3(pteval_t, mmu.ptep_modify_prot_start, vma, addr, ptep);
+	ret = PVOP_CALL3(pteval_t, pv_ops, mmu.ptep_modify_prot_start, vma, addr, ptep);
 
 	return (pte_t) { .pte = ret };
 }
@@ -380,41 +380,41 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, unsigned
 					   pte_t *ptep, pte_t old_pte, pte_t pte)
 {
 
-	PVOP_VCALL4(mmu.ptep_modify_prot_commit, vma, addr, ptep, pte.pte);
+	PVOP_VCALL4(pv_ops, mmu.ptep_modify_prot_commit, vma, addr, ptep, pte.pte);
 }
 
 static inline void set_pte(pte_t *ptep, pte_t pte)
 {
-	PVOP_VCALL2(mmu.set_pte, ptep, pte.pte);
+	PVOP_VCALL2(pv_ops, mmu.set_pte, ptep, pte.pte);
 }
 
 static inline void set_pmd(pmd_t *pmdp, pmd_t pmd)
 {
-	PVOP_VCALL2(mmu.set_pmd, pmdp, native_pmd_val(pmd));
+	PVOP_VCALL2(pv_ops, mmu.set_pmd, pmdp, native_pmd_val(pmd));
 }
 
 static inline pmd_t __pmd(pmdval_t val)
 {
-	return (pmd_t) { PVOP_ALT_CALLEE1(pmdval_t, mmu.make_pmd, val,
+	return (pmd_t) { PVOP_ALT_CALLEE1(pmdval_t, pv_ops, mmu.make_pmd, val,
 					  "mov %%rdi, %%rax", ALT_NOT_XEN) };
 }
 
 static inline pmdval_t pmd_val(pmd_t pmd)
 {
-	return PVOP_ALT_CALLEE1(pmdval_t, mmu.pmd_val, pmd.pmd,
+	return PVOP_ALT_CALLEE1(pmdval_t, pv_ops, mmu.pmd_val, pmd.pmd,
 				"mov %%rdi, %%rax", ALT_NOT_XEN);
 }
 
 static inline void set_pud(pud_t *pudp, pud_t pud)
 {
-	PVOP_VCALL2(mmu.set_pud, pudp, native_pud_val(pud));
+	PVOP_VCALL2(pv_ops, mmu.set_pud, pudp, native_pud_val(pud));
 }
 
 static inline pud_t __pud(pudval_t val)
 {
 	pudval_t ret;
 
-	ret = PVOP_ALT_CALLEE1(pudval_t, mmu.make_pud, val,
+	ret = PVOP_ALT_CALLEE1(pudval_t, pv_ops, mmu.make_pud, val,
 			       "mov %%rdi, %%rax", ALT_NOT_XEN);
 
 	return (pud_t) { ret };
@@ -422,7 +422,7 @@ static inline pud_t __pud(pudval_t val)
 
 static inline pudval_t pud_val(pud_t pud)
 {
-	return PVOP_ALT_CALLEE1(pudval_t, mmu.pud_val, pud.pud,
+	return PVOP_ALT_CALLEE1(pudval_t, pv_ops, mmu.pud_val, pud.pud,
 				"mov %%rdi, %%rax", ALT_NOT_XEN);
 }
 
@@ -435,12 +435,12 @@ static inline void set_p4d(p4d_t *p4dp, p4d_t p4d)
 {
 	p4dval_t val = native_p4d_val(p4d);
 
-	PVOP_VCALL2(mmu.set_p4d, p4dp, val);
+	PVOP_VCALL2(pv_ops, mmu.set_p4d, p4dp, val);
 }
 
 static inline p4d_t __p4d(p4dval_t val)
 {
-	p4dval_t ret = PVOP_ALT_CALLEE1(p4dval_t, mmu.make_p4d, val,
+	p4dval_t ret = PVOP_ALT_CALLEE1(p4dval_t, pv_ops, mmu.make_p4d, val,
 					"mov %%rdi, %%rax", ALT_NOT_XEN);
 
 	return (p4d_t) { ret };
@@ -448,13 +448,13 @@ static inline p4d_t __p4d(p4dval_t val)
 
 static inline p4dval_t p4d_val(p4d_t p4d)
 {
-	return PVOP_ALT_CALLEE1(p4dval_t, mmu.p4d_val, p4d.p4d,
+	return PVOP_ALT_CALLEE1(p4dval_t, pv_ops, mmu.p4d_val, p4d.p4d,
 				"mov %%rdi, %%rax", ALT_NOT_XEN);
 }
 
 static inline void __set_pgd(pgd_t *pgdp, pgd_t pgd)
 {
-	PVOP_VCALL2(mmu.set_pgd, pgdp, native_pgd_val(pgd));
+	PVOP_VCALL2(pv_ops, mmu.set_pgd, pgdp, native_pgd_val(pgd));
 }
 
 #define set_pgd(pgdp, pgdval) do {					\
@@ -493,28 +493,28 @@ static inline void pmd_clear(pmd_t *pmdp)
 #define  __HAVE_ARCH_START_CONTEXT_SWITCH
 static inline void arch_start_context_switch(struct task_struct *prev)
 {
-	PVOP_VCALL1(cpu.start_context_switch, prev);
+	PVOP_VCALL1(pv_ops, cpu.start_context_switch, prev);
 }
 
 static inline void arch_end_context_switch(struct task_struct *next)
 {
-	PVOP_VCALL1(cpu.end_context_switch, next);
+	PVOP_VCALL1(pv_ops, cpu.end_context_switch, next);
 }
 
 #define  __HAVE_ARCH_ENTER_LAZY_MMU_MODE
 static inline void arch_enter_lazy_mmu_mode(void)
 {
-	PVOP_VCALL0(mmu.lazy_mode.enter);
+	PVOP_VCALL0(pv_ops, mmu.lazy_mode.enter);
 }
 
 static inline void arch_leave_lazy_mmu_mode(void)
 {
-	PVOP_VCALL0(mmu.lazy_mode.leave);
+	PVOP_VCALL0(pv_ops, mmu.lazy_mode.leave);
 }
 
 static inline void arch_flush_lazy_mmu_mode(void)
 {
-	PVOP_VCALL0(mmu.lazy_mode.flush);
+	PVOP_VCALL0(pv_ops, mmu.lazy_mode.flush);
 }
 
 static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx,
@@ -529,29 +529,29 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx,
 static __always_inline void pv_queued_spin_lock_slowpath(struct qspinlock *lock,
 							u32 val)
 {
-	PVOP_VCALL2(lock.queued_spin_lock_slowpath, lock, val);
+	PVOP_VCALL2(pv_ops, lock.queued_spin_lock_slowpath, lock, val);
 }
 
 static __always_inline void pv_queued_spin_unlock(struct qspinlock *lock)
 {
-	PVOP_ALT_VCALLEE1(lock.queued_spin_unlock, lock,
+	PVOP_ALT_VCALLEE1(pv_ops, lock.queued_spin_unlock, lock,
 			  "movb $0, (%%" _ASM_ARG1 ");",
 			  ALT_NOT(X86_FEATURE_PVUNLOCK));
 }
 
 static __always_inline void pv_wait(u8 *ptr, u8 val)
 {
-	PVOP_VCALL2(lock.wait, ptr, val);
+	PVOP_VCALL2(pv_ops, lock.wait, ptr, val);
 }
 
 static __always_inline void pv_kick(int cpu)
 {
-	PVOP_VCALL1(lock.kick, cpu);
+	PVOP_VCALL1(pv_ops, lock.kick, cpu);
 }
 
 static __always_inline bool pv_vcpu_is_preempted(long cpu)
 {
-	return PVOP_ALT_CALLEE1(bool, lock.vcpu_is_preempted, cpu,
+	return PVOP_ALT_CALLEE1(bool, pv_ops, lock.vcpu_is_preempted, cpu,
 				"xor %%" _ASM_AX ", %%" _ASM_AX ";",
 				ALT_NOT(X86_FEATURE_VCPUPREEMPT));
 }
@@ -564,18 +564,18 @@ bool __raw_callee_save___native_vcpu_is_preempted(long cpu);
 #ifdef CONFIG_PARAVIRT_XXL
 static __always_inline unsigned long arch_local_save_flags(void)
 {
-	return PVOP_ALT_CALLEE0(unsigned long, irq.save_fl, "pushf; pop %%rax;",
+	return PVOP_ALT_CALLEE0(unsigned long, pv_ops, irq.save_fl, "pushf; pop %%rax;",
 				ALT_NOT_XEN);
 }
 
 static __always_inline void arch_local_irq_disable(void)
 {
-	PVOP_ALT_VCALLEE0(irq.irq_disable, "cli;", ALT_NOT_XEN);
+	PVOP_ALT_VCALLEE0(pv_ops, irq.irq_disable, "cli;", ALT_NOT_XEN);
 }
 
 static __always_inline void arch_local_irq_enable(void)
 {
-	PVOP_ALT_VCALLEE0(irq.irq_enable, "sti;", ALT_NOT_XEN);
+	PVOP_ALT_VCALLEE0(pv_ops, irq.irq_enable, "sti;", ALT_NOT_XEN);
 }
 
 static __always_inline unsigned long arch_local_irq_save(void)
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 1e50f13e6543..01a485f1a7f1 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -212,7 +212,7 @@ struct paravirt_patch_template {
 
 extern struct paravirt_patch_template pv_ops;
 
-#define paravirt_ptr(op)	[paravirt_opptr] "m" (pv_ops.op)
+#define paravirt_ptr(array, op)	[paravirt_opptr] "m" (array.op)
 
 /*
  * This generates an indirect call based on the operation type number.
@@ -362,19 +362,19 @@ extern struct paravirt_patch_template pv_ops;
  *   feature is not active, the direct call is used as above via the
  *   ALT_FLAG_DIRECT_CALL special case and the "always on" feature.
  */
-#define ____PVOP_CALL(ret, op, call_clbr, extra_clbr, ...)	\
+#define ____PVOP_CALL(ret, array, op, call_clbr, extra_clbr, ...)	\
 	({								\
 		PVOP_CALL_ARGS;						\
 		asm volatile(ALTERNATIVE(PARAVIRT_CALL, ALT_CALL_INSTR,	\
 				ALT_CALL_ALWAYS)			\
 			     : call_clbr, ASM_CALL_CONSTRAINT		\
-			     : paravirt_ptr(op),			\
+			     : paravirt_ptr(array, op),			\
 			       ##__VA_ARGS__				\
 			     : "memory", "cc" extra_clbr);		\
 		ret;							\
 	})
 
-#define ____PVOP_ALT_CALL(ret, op, alt, cond, call_clbr,		\
+#define ____PVOP_ALT_CALL(ret, array, op, alt, cond, call_clbr,		\
 			  extra_clbr, ...)				\
 	({								\
 		PVOP_CALL_ARGS;						\
@@ -382,102 +382,102 @@ extern struct paravirt_patch_template pv_ops;
 				 ALT_CALL_INSTR, ALT_CALL_ALWAYS,	\
 				 alt, cond)				\
 			     : call_clbr, ASM_CALL_CONSTRAINT		\
-			     : paravirt_ptr(op),			\
+			     : paravirt_ptr(array, op),			\
 			       ##__VA_ARGS__				\
 			     : "memory", "cc" extra_clbr);		\
 		ret;							\
 	})
 
-#define __PVOP_CALL(rettype, op, ...)					\
-	____PVOP_CALL(PVOP_RETVAL(rettype), op,				\
+#define __PVOP_CALL(rettype, array, op, ...)				\
+	____PVOP_CALL(PVOP_RETVAL(rettype), array, op,			\
 		      PVOP_CALL_CLOBBERS, EXTRA_CLOBBERS, ##__VA_ARGS__)
 
-#define __PVOP_ALT_CALL(rettype, op, alt, cond, ...)			\
-	____PVOP_ALT_CALL(PVOP_RETVAL(rettype), op, alt, cond,		\
+#define __PVOP_ALT_CALL(rettype, array, op, alt, cond, ...)		\
+	____PVOP_ALT_CALL(PVOP_RETVAL(rettype), array, op, alt, cond,	\
 			  PVOP_CALL_CLOBBERS, EXTRA_CLOBBERS,		\
 			  ##__VA_ARGS__)
 
-#define __PVOP_CALLEESAVE(rettype, op, ...)				\
-	____PVOP_CALL(PVOP_RETVAL(rettype), op.func,			\
+#define __PVOP_CALLEESAVE(rettype, array, op, ...)			\
+	____PVOP_CALL(PVOP_RETVAL(rettype), array, op.func,		\
 		      PVOP_CALLEE_CLOBBERS, , ##__VA_ARGS__)
 
-#define __PVOP_ALT_CALLEESAVE(rettype, op, alt, cond, ...)		\
-	____PVOP_ALT_CALL(PVOP_RETVAL(rettype), op.func, alt, cond,	\
+#define __PVOP_ALT_CALLEESAVE(rettype, array, op, alt, cond, ...)	\
+	____PVOP_ALT_CALL(PVOP_RETVAL(rettype), array, op.func, alt, cond, \
 			  PVOP_CALLEE_CLOBBERS, , ##__VA_ARGS__)
 
 
-#define __PVOP_VCALL(op, ...)						\
-	(void)____PVOP_CALL(, op, PVOP_VCALL_CLOBBERS,			\
+#define __PVOP_VCALL(array, op, ...)					\
+	(void)____PVOP_CALL(, array, op, PVOP_VCALL_CLOBBERS,		\
 		       VEXTRA_CLOBBERS, ##__VA_ARGS__)
 
-#define __PVOP_ALT_VCALL(op, alt, cond, ...)				\
-	(void)____PVOP_ALT_CALL(, op, alt, cond,			\
+#define __PVOP_ALT_VCALL(array, op, alt, cond, ...)			\
+	(void)____PVOP_ALT_CALL(, array, op, alt, cond,			\
 				PVOP_VCALL_CLOBBERS, VEXTRA_CLOBBERS,	\
 				##__VA_ARGS__)
 
-#define __PVOP_VCALLEESAVE(op, ...)					\
-	(void)____PVOP_CALL(, op.func,					\
+#define __PVOP_VCALLEESAVE(array, op, ...)				\
+	(void)____PVOP_CALL(, array, op.func,				\
 			    PVOP_VCALLEE_CLOBBERS, , ##__VA_ARGS__)
 
-#define __PVOP_ALT_VCALLEESAVE(op, alt, cond, ...)			\
-	(void)____PVOP_ALT_CALL(, op.func, alt, cond,			\
+#define __PVOP_ALT_VCALLEESAVE(array, op, alt, cond, ...)		\
+	(void)____PVOP_ALT_CALL(, array, op.func, alt, cond,		\
 				PVOP_VCALLEE_CLOBBERS, , ##__VA_ARGS__)
 
 
-#define PVOP_CALL0(rettype, op)						\
-	__PVOP_CALL(rettype, op)
-#define PVOP_VCALL0(op)							\
-	__PVOP_VCALL(op)
-#define PVOP_ALT_CALL0(rettype, op, alt, cond)				\
-	__PVOP_ALT_CALL(rettype, op, alt, cond)
-#define PVOP_ALT_VCALL0(op, alt, cond)					\
-	__PVOP_ALT_VCALL(op, alt, cond)
-
-#define PVOP_CALLEE0(rettype, op)					\
-	__PVOP_CALLEESAVE(rettype, op)
-#define PVOP_VCALLEE0(op)						\
-	__PVOP_VCALLEESAVE(op)
-#define PVOP_ALT_CALLEE0(rettype, op, alt, cond)			\
-	__PVOP_ALT_CALLEESAVE(rettype, op, alt, cond)
-#define PVOP_ALT_VCALLEE0(op, alt, cond)				\
-	__PVOP_ALT_VCALLEESAVE(op, alt, cond)
-
-
-#define PVOP_CALL1(rettype, op, arg1)					\
-	__PVOP_CALL(rettype, op, PVOP_CALL_ARG1(arg1))
-#define PVOP_VCALL1(op, arg1)						\
-	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1))
-#define PVOP_ALT_VCALL1(op, arg1, alt, cond)				\
-	__PVOP_ALT_VCALL(op, alt, cond, PVOP_CALL_ARG1(arg1))
-
-#define PVOP_CALLEE1(rettype, op, arg1)					\
-	__PVOP_CALLEESAVE(rettype, op, PVOP_CALL_ARG1(arg1))
-#define PVOP_VCALLEE1(op, arg1)						\
-	__PVOP_VCALLEESAVE(op, PVOP_CALL_ARG1(arg1))
-#define PVOP_ALT_CALLEE1(rettype, op, arg1, alt, cond)			\
-	__PVOP_ALT_CALLEESAVE(rettype, op, alt, cond, PVOP_CALL_ARG1(arg1))
-#define PVOP_ALT_VCALLEE1(op, arg1, alt, cond)				\
-	__PVOP_ALT_VCALLEESAVE(op, alt, cond, PVOP_CALL_ARG1(arg1))
-
-
-#define PVOP_CALL2(rettype, op, arg1, arg2)				\
-	__PVOP_CALL(rettype, op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2))
-#define PVOP_VCALL2(op, arg1, arg2)					\
-	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2))
-
-#define PVOP_CALL3(rettype, op, arg1, arg2, arg3)			\
-	__PVOP_CALL(rettype, op, PVOP_CALL_ARG1(arg1),			\
+#define PVOP_CALL0(rettype, array, op)					\
+	__PVOP_CALL(rettype, array, op)
+#define PVOP_VCALL0(array, op)						\
+	__PVOP_VCALL(array, op)
+#define PVOP_ALT_CALL0(rettype, array, op, alt, cond)			\
+	__PVOP_ALT_CALL(rettype, array, op, alt, cond)
+#define PVOP_ALT_VCALL0(array, op, alt, cond)				\
+	__PVOP_ALT_VCALL(array, op, alt, cond)
+
+#define PVOP_CALLEE0(rettype, array, op)				\
+	__PVOP_CALLEESAVE(rettype, array, op)
+#define PVOP_VCALLEE0(array, op)					\
+	__PVOP_VCALLEESAVE(array, op)
+#define PVOP_ALT_CALLEE0(rettype, array, op, alt, cond)			\
+	__PVOP_ALT_CALLEESAVE(rettype, array, op, alt, cond)
+#define PVOP_ALT_VCALLEE0(array, op, alt, cond)				\
+	__PVOP_ALT_VCALLEESAVE(array, op, alt, cond)
+
+
+#define PVOP_CALL1(rettype, array, op, arg1)				\
+	__PVOP_CALL(rettype, array, op, PVOP_CALL_ARG1(arg1))
+#define PVOP_VCALL1(array, op, arg1)					\
+	__PVOP_VCALL(array, op, PVOP_CALL_ARG1(arg1))
+#define PVOP_ALT_VCALL1(array, op, arg1, alt, cond)			\
+	__PVOP_ALT_VCALL(array, op, alt, cond, PVOP_CALL_ARG1(arg1))
+
+#define PVOP_CALLEE1(rettype, array, op, arg1)				\
+	__PVOP_CALLEESAVE(rettype, array, op, PVOP_CALL_ARG1(arg1))
+#define PVOP_VCALLEE1(array, op, arg1)					\
+	__PVOP_VCALLEESAVE(array, op, PVOP_CALL_ARG1(arg1))
+#define PVOP_ALT_CALLEE1(rettype, array, op, arg1, alt, cond)		\
+	__PVOP_ALT_CALLEESAVE(rettype, array, op, alt, cond, PVOP_CALL_ARG1(arg1))
+#define PVOP_ALT_VCALLEE1(array, op, arg1, alt, cond)			\
+	__PVOP_ALT_VCALLEESAVE(array, op, alt, cond, PVOP_CALL_ARG1(arg1))
+
+
+#define PVOP_CALL2(rettype, array, op, arg1, arg2)			\
+	__PVOP_CALL(rettype, array, op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2))
+#define PVOP_VCALL2(array, op, arg1, arg2)				\
+	__PVOP_VCALL(array, op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2))
+
+#define PVOP_CALL3(rettype, array, op, arg1, arg2, arg3)		\
+	__PVOP_CALL(rettype, array, op, PVOP_CALL_ARG1(arg1),		\
 		    PVOP_CALL_ARG2(arg2), PVOP_CALL_ARG3(arg3))
-#define PVOP_VCALL3(op, arg1, arg2, arg3)				\
-	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1),				\
+#define PVOP_VCALL3(array, op, arg1, arg2, arg3)			\
+	__PVOP_VCALL(array, op, PVOP_CALL_ARG1(arg1),			\
 		     PVOP_CALL_ARG2(arg2), PVOP_CALL_ARG3(arg3))
 
-#define PVOP_CALL4(rettype, op, arg1, arg2, arg3, arg4)			\
-	__PVOP_CALL(rettype, op,					\
+#define PVOP_CALL4(rettype, array, op, arg1, arg2, arg3, arg4)		\
+	__PVOP_CALL(rettype, array, op,					\
 		    PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2),		\
 		    PVOP_CALL_ARG3(arg3), PVOP_CALL_ARG4(arg4))
-#define PVOP_VCALL4(op, arg1, arg2, arg3, arg4)				\
-	__PVOP_VCALL(op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2),	\
+#define PVOP_VCALL4(array, op, arg1, arg2, arg3, arg4)			\
+	__PVOP_VCALL(array, op, PVOP_CALL_ARG1(arg1), PVOP_CALL_ARG2(arg2), \
 		     PVOP_CALL_ARG3(arg3), PVOP_CALL_ARG4(arg4))
 
 #endif	/* __ASSEMBLER__ */
-- 
2.51.0

[PATCH v4 21/21] x86/pvlocks: Move paravirt spinlock functions into own header

[Juergen Gross]

Instead of having the pv spinlock function definitions in paravirt.h,
move them into the new header paravirt-spinlock.h.

Signed-off-by: Juergen Gross <jgross@suse.com>
---
V2:
- use new header instead of qspinlock.h
- use dedicated pv_ops_lock array
- move more paravirt related lock code
V3:
- hide native_pv_lock_init() with CONFIG_SMP (kernel test robot)
V4:
- don't reference pv_ops_lock without CONFIG_PARAVIRT_SPINLOCKS
  (kernel test robot)
---
 arch/x86/hyperv/hv_spinlock.c            |  10 +-
 arch/x86/include/asm/paravirt-spinlock.h | 146 +++++++++++++++++++++++
 arch/x86/include/asm/paravirt.h          |  61 ----------
 arch/x86/include/asm/paravirt_types.h    |  17 ---
 arch/x86/include/asm/qspinlock.h         |  89 ++------------
 arch/x86/kernel/Makefile                 |   2 +-
 arch/x86/kernel/kvm.c                    |  12 +-
 arch/x86/kernel/paravirt-spinlocks.c     |  26 +++-
 arch/x86/kernel/paravirt.c               |  21 ----
 arch/x86/xen/spinlock.c                  |  10 +-
 tools/objtool/check.c                    |   1 +
 11 files changed, 196 insertions(+), 199 deletions(-)
 create mode 100644 arch/x86/include/asm/paravirt-spinlock.h

diff --git a/arch/x86/hyperv/hv_spinlock.c b/arch/x86/hyperv/hv_spinlock.c
index 2a3c2afb0154..210b494e4de0 100644
--- a/arch/x86/hyperv/hv_spinlock.c
+++ b/arch/x86/hyperv/hv_spinlock.c
@@ -78,11 +78,11 @@ void __init hv_init_spinlocks(void)
 	pr_info("PV spinlocks enabled\n");
 
 	__pv_init_lock_hash();
-	pv_ops.lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
-	pv_ops.lock.queued_spin_unlock = PV_CALLEE_SAVE(__pv_queued_spin_unlock);
-	pv_ops.lock.wait = hv_qlock_wait;
-	pv_ops.lock.kick = hv_qlock_kick;
-	pv_ops.lock.vcpu_is_preempted = PV_CALLEE_SAVE(hv_vcpu_is_preempted);
+	pv_ops_lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
+	pv_ops_lock.queued_spin_unlock = PV_CALLEE_SAVE(__pv_queued_spin_unlock);
+	pv_ops_lock.wait = hv_qlock_wait;
+	pv_ops_lock.kick = hv_qlock_kick;
+	pv_ops_lock.vcpu_is_preempted = PV_CALLEE_SAVE(hv_vcpu_is_preempted);
 }
 
 static __init int hv_parse_nopvspin(char *arg)
diff --git a/arch/x86/include/asm/paravirt-spinlock.h b/arch/x86/include/asm/paravirt-spinlock.h
new file mode 100644
index 000000000000..ed3ed343903d
--- /dev/null
+++ b/arch/x86/include/asm/paravirt-spinlock.h
@@ -0,0 +1,146 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef _ASM_X86_PARAVIRT_SPINLOCK_H
+#define _ASM_X86_PARAVIRT_SPINLOCK_H
+
+#include <asm/paravirt_types.h>
+
+#ifdef CONFIG_SMP
+#include <asm/spinlock_types.h>
+#endif
+
+struct qspinlock;
+
+struct pv_lock_ops {
+	void (*queued_spin_lock_slowpath)(struct qspinlock *lock, u32 val);
+	struct paravirt_callee_save queued_spin_unlock;
+
+	void (*wait)(u8 *ptr, u8 val);
+	void (*kick)(int cpu);
+
+	struct paravirt_callee_save vcpu_is_preempted;
+} __no_randomize_layout;
+
+extern struct pv_lock_ops pv_ops_lock;
+
+#ifdef CONFIG_PARAVIRT_SPINLOCKS
+void __init paravirt_set_cap(void);
+extern void native_queued_spin_lock_slowpath(struct qspinlock *lock, u32 val);
+extern void __pv_init_lock_hash(void);
+extern void __pv_queued_spin_lock_slowpath(struct qspinlock *lock, u32 val);
+extern void __raw_callee_save___pv_queued_spin_unlock(struct qspinlock *lock);
+extern bool nopvspin;
+
+static __always_inline void pv_queued_spin_lock_slowpath(struct qspinlock *lock,
+							 u32 val)
+{
+	PVOP_VCALL2(pv_ops_lock, queued_spin_lock_slowpath, lock, val);
+}
+
+static __always_inline void pv_queued_spin_unlock(struct qspinlock *lock)
+{
+	PVOP_ALT_VCALLEE1(pv_ops_lock, queued_spin_unlock, lock,
+			  "movb $0, (%%" _ASM_ARG1 ");",
+			  ALT_NOT(X86_FEATURE_PVUNLOCK));
+}
+
+static __always_inline bool pv_vcpu_is_preempted(long cpu)
+{
+	return PVOP_ALT_CALLEE1(bool, pv_ops_lock, vcpu_is_preempted, cpu,
+				"xor %%" _ASM_AX ", %%" _ASM_AX ";",
+				ALT_NOT(X86_FEATURE_VCPUPREEMPT));
+}
+
+#define queued_spin_unlock queued_spin_unlock
+/**
+ * queued_spin_unlock - release a queued spinlock
+ * @lock : Pointer to queued spinlock structure
+ *
+ * A smp_store_release() on the least-significant byte.
+ */
+static inline void native_queued_spin_unlock(struct qspinlock *lock)
+{
+	smp_store_release(&lock->locked, 0);
+}
+
+static inline void queued_spin_lock_slowpath(struct qspinlock *lock, u32 val)
+{
+	pv_queued_spin_lock_slowpath(lock, val);
+}
+
+static inline void queued_spin_unlock(struct qspinlock *lock)
+{
+	kcsan_release();
+	pv_queued_spin_unlock(lock);
+}
+
+#define vcpu_is_preempted vcpu_is_preempted
+static inline bool vcpu_is_preempted(long cpu)
+{
+	return pv_vcpu_is_preempted(cpu);
+}
+
+static __always_inline void pv_wait(u8 *ptr, u8 val)
+{
+	PVOP_VCALL2(pv_ops_lock, wait, ptr, val);
+}
+
+static __always_inline void pv_kick(int cpu)
+{
+	PVOP_VCALL1(pv_ops_lock, kick, cpu);
+}
+
+void __raw_callee_save___native_queued_spin_unlock(struct qspinlock *lock);
+bool __raw_callee_save___native_vcpu_is_preempted(long cpu);
+#endif /* CONFIG_PARAVIRT_SPINLOCKS */
+
+void __init native_pv_lock_init(void);
+__visible void __native_queued_spin_unlock(struct qspinlock *lock);
+bool pv_is_native_spin_unlock(void);
+__visible bool __native_vcpu_is_preempted(long cpu);
+bool pv_is_native_vcpu_is_preempted(void);
+
+/*
+ * virt_spin_lock_key - disables by default the virt_spin_lock() hijack.
+ *
+ * Native (and PV wanting native due to vCPU pinning) should keep this key
+ * disabled. Native does not touch the key.
+ *
+ * When in a guest then native_pv_lock_init() enables the key first and
+ * KVM/XEN might conditionally disable it later in the boot process again.
+ */
+DECLARE_STATIC_KEY_FALSE(virt_spin_lock_key);
+
+/*
+ * Shortcut for the queued_spin_lock_slowpath() function that allows
+ * virt to hijack it.
+ *
+ * Returns:
+ *   true - lock has been negotiated, all done;
+ *   false - queued_spin_lock_slowpath() will do its thing.
+ */
+#define virt_spin_lock virt_spin_lock
+static inline bool virt_spin_lock(struct qspinlock *lock)
+{
+	int val;
+
+	if (!static_branch_likely(&virt_spin_lock_key))
+		return false;
+
+	/*
+	 * On hypervisors without PARAVIRT_SPINLOCKS support we fall
+	 * back to a Test-and-Set spinlock, because fair locks have
+	 * horrible lock 'holder' preemption issues.
+	 */
+
+ __retry:
+	val = atomic_read(&lock->val);
+
+	if (val || !atomic_try_cmpxchg(&lock->val, &val, _Q_LOCKED_VAL)) {
+		cpu_relax();
+		goto __retry;
+	}
+
+	return true;
+}
+
+#endif /* _ASM_X86_PARAVIRT_SPINLOCK_H */
diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h
index ec274d13bae0..b21072af731d 100644
--- a/arch/x86/include/asm/paravirt.h
+++ b/arch/x86/include/asm/paravirt.h
@@ -19,15 +19,6 @@
 #include <linux/cpumask.h>
 #include <asm/frame.h>
 
-__visible void __native_queued_spin_unlock(struct qspinlock *lock);
-bool pv_is_native_spin_unlock(void);
-__visible bool __native_vcpu_is_preempted(long cpu);
-bool pv_is_native_vcpu_is_preempted(void);
-
-#ifdef CONFIG_PARAVIRT_SPINLOCKS
-void __init paravirt_set_cap(void);
-#endif
-
 /* The paravirtualized I/O functions */
 static inline void slow_down_io(void)
 {
@@ -522,46 +513,7 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx,
 {
 	pv_ops.mmu.set_fixmap(idx, phys, flags);
 }
-#endif
-
-#if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS)
-
-static __always_inline void pv_queued_spin_lock_slowpath(struct qspinlock *lock,
-							u32 val)
-{
-	PVOP_VCALL2(pv_ops, lock.queued_spin_lock_slowpath, lock, val);
-}
-
-static __always_inline void pv_queued_spin_unlock(struct qspinlock *lock)
-{
-	PVOP_ALT_VCALLEE1(pv_ops, lock.queued_spin_unlock, lock,
-			  "movb $0, (%%" _ASM_ARG1 ");",
-			  ALT_NOT(X86_FEATURE_PVUNLOCK));
-}
-
-static __always_inline void pv_wait(u8 *ptr, u8 val)
-{
-	PVOP_VCALL2(pv_ops, lock.wait, ptr, val);
-}
-
-static __always_inline void pv_kick(int cpu)
-{
-	PVOP_VCALL1(pv_ops, lock.kick, cpu);
-}
-
-static __always_inline bool pv_vcpu_is_preempted(long cpu)
-{
-	return PVOP_ALT_CALLEE1(bool, pv_ops, lock.vcpu_is_preempted, cpu,
-				"xor %%" _ASM_AX ", %%" _ASM_AX ";",
-				ALT_NOT(X86_FEATURE_VCPUPREEMPT));
-}
 
-void __raw_callee_save___native_queued_spin_unlock(struct qspinlock *lock);
-bool __raw_callee_save___native_vcpu_is_preempted(long cpu);
-
-#endif /* SMP && PARAVIRT_SPINLOCKS */
-
-#ifdef CONFIG_PARAVIRT_XXL
 static __always_inline unsigned long arch_local_save_flags(void)
 {
 	return PVOP_ALT_CALLEE0(unsigned long, pv_ops, irq.save_fl, "pushf; pop %%rax;",
@@ -588,8 +540,6 @@ static __always_inline unsigned long arch_local_irq_save(void)
 }
 #endif
 
-void native_pv_lock_init(void) __init;
-
 #else  /* __ASSEMBLER__ */
 
 #ifdef CONFIG_X86_64
@@ -613,12 +563,6 @@ void native_pv_lock_init(void) __init;
 #endif /* __ASSEMBLER__ */
 #else  /* CONFIG_PARAVIRT */
 # define default_banner x86_init_noop
-
-#ifndef __ASSEMBLER__
-static inline void native_pv_lock_init(void)
-{
-}
-#endif
 #endif /* !CONFIG_PARAVIRT */
 
 #ifndef __ASSEMBLER__
@@ -634,10 +578,5 @@ static inline void paravirt_arch_exit_mmap(struct mm_struct *mm)
 }
 #endif
 
-#ifndef CONFIG_PARAVIRT_SPINLOCKS
-static inline void paravirt_set_cap(void)
-{
-}
-#endif
 #endif /* __ASSEMBLER__ */
 #endif /* _ASM_X86_PARAVIRT_H */
diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h
index 01a485f1a7f1..e2b487d35d14 100644
--- a/arch/x86/include/asm/paravirt_types.h
+++ b/arch/x86/include/asm/paravirt_types.h
@@ -184,22 +184,6 @@ struct pv_mmu_ops {
 #endif
 } __no_randomize_layout;
 
-#ifdef CONFIG_SMP
-#include <asm/spinlock_types.h>
-#endif
-
-struct qspinlock;
-
-struct pv_lock_ops {
-	void (*queued_spin_lock_slowpath)(struct qspinlock *lock, u32 val);
-	struct paravirt_callee_save queued_spin_unlock;
-
-	void (*wait)(u8 *ptr, u8 val);
-	void (*kick)(int cpu);
-
-	struct paravirt_callee_save vcpu_is_preempted;
-} __no_randomize_layout;
-
 /* This contains all the paravirt structures: we get a convenient
  * number for each function using the offset which we use to indicate
  * what to patch. */
@@ -207,7 +191,6 @@ struct paravirt_patch_template {
 	struct pv_cpu_ops	cpu;
 	struct pv_irq_ops	irq;
 	struct pv_mmu_ops	mmu;
-	struct pv_lock_ops	lock;
 } __no_randomize_layout;
 
 extern struct paravirt_patch_template pv_ops;
diff --git a/arch/x86/include/asm/qspinlock.h b/arch/x86/include/asm/qspinlock.h
index 68da67df304d..a2668bdf4c84 100644
--- a/arch/x86/include/asm/qspinlock.h
+++ b/arch/x86/include/asm/qspinlock.h
@@ -7,6 +7,9 @@
 #include <asm-generic/qspinlock_types.h>
 #include <asm/paravirt.h>
 #include <asm/rmwcc.h>
+#ifdef CONFIG_PARAVIRT
+#include <asm/paravirt-spinlock.h>
+#endif
 
 #define _Q_PENDING_LOOPS	(1 << 9)
 
@@ -27,89 +30,13 @@ static __always_inline u32 queued_fetch_set_pending_acquire(struct qspinlock *lo
 	return val;
 }
 
-#ifdef CONFIG_PARAVIRT_SPINLOCKS
-extern void native_queued_spin_lock_slowpath(struct qspinlock *lock, u32 val);
-extern void __pv_init_lock_hash(void);
-extern void __pv_queued_spin_lock_slowpath(struct qspinlock *lock, u32 val);
-extern void __raw_callee_save___pv_queued_spin_unlock(struct qspinlock *lock);
-extern bool nopvspin;
-
-#define	queued_spin_unlock queued_spin_unlock
-/**
- * queued_spin_unlock - release a queued spinlock
- * @lock : Pointer to queued spinlock structure
- *
- * A smp_store_release() on the least-significant byte.
- */
-static inline void native_queued_spin_unlock(struct qspinlock *lock)
-{
-	smp_store_release(&lock->locked, 0);
-}
-
-static inline void queued_spin_lock_slowpath(struct qspinlock *lock, u32 val)
-{
-	pv_queued_spin_lock_slowpath(lock, val);
-}
-
-static inline void queued_spin_unlock(struct qspinlock *lock)
-{
-	kcsan_release();
-	pv_queued_spin_unlock(lock);
-}
-
-#define vcpu_is_preempted vcpu_is_preempted
-static inline bool vcpu_is_preempted(long cpu)
-{
-	return pv_vcpu_is_preempted(cpu);
-}
+#ifndef CONFIG_PARAVIRT_SPINLOCKS
+static inline void paravirt_set_cap(void) { }
 #endif
 
-#ifdef CONFIG_PARAVIRT
-/*
- * virt_spin_lock_key - disables by default the virt_spin_lock() hijack.
- *
- * Native (and PV wanting native due to vCPU pinning) should keep this key
- * disabled. Native does not touch the key.
- *
- * When in a guest then native_pv_lock_init() enables the key first and
- * KVM/XEN might conditionally disable it later in the boot process again.
- */
-DECLARE_STATIC_KEY_FALSE(virt_spin_lock_key);
-
-/*
- * Shortcut for the queued_spin_lock_slowpath() function that allows
- * virt to hijack it.
- *
- * Returns:
- *   true - lock has been negotiated, all done;
- *   false - queued_spin_lock_slowpath() will do its thing.
- */
-#define virt_spin_lock virt_spin_lock
-static inline bool virt_spin_lock(struct qspinlock *lock)
-{
-	int val;
-
-	if (!static_branch_likely(&virt_spin_lock_key))
-		return false;
-
-	/*
-	 * On hypervisors without PARAVIRT_SPINLOCKS support we fall
-	 * back to a Test-and-Set spinlock, because fair locks have
-	 * horrible lock 'holder' preemption issues.
-	 */
-
- __retry:
-	val = atomic_read(&lock->val);
-
-	if (val || !atomic_try_cmpxchg(&lock->val, &val, _Q_LOCKED_VAL)) {
-		cpu_relax();
-		goto __retry;
-	}
-
-	return true;
-}
-
-#endif /* CONFIG_PARAVIRT */
+#ifndef CONFIG_PARAVIRT
+static inline void native_pv_lock_init(void) { }
+#endif
 
 #include <asm-generic/qspinlock.h>
 
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
index bc184dd38d99..e9aeeeafad17 100644
--- a/arch/x86/kernel/Makefile
+++ b/arch/x86/kernel/Makefile
@@ -126,7 +126,7 @@ obj-$(CONFIG_DEBUG_NMI_SELFTEST) += nmi_selftest.o
 
 obj-$(CONFIG_KVM_GUEST)		+= kvm.o kvmclock.o
 obj-$(CONFIG_PARAVIRT)		+= paravirt.o
-obj-$(CONFIG_PARAVIRT_SPINLOCKS)+= paravirt-spinlocks.o
+obj-$(CONFIG_PARAVIRT)		+= paravirt-spinlocks.o
 obj-$(CONFIG_PARAVIRT_CLOCK)	+= pvclock.o
 obj-$(CONFIG_X86_PMEM_LEGACY_DEVICE) += pmem.o
 
diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index d54fd2bc0402..e767f8ed405a 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -824,8 +824,10 @@ static void __init kvm_guest_init(void)
 		has_steal_clock = 1;
 		static_call_update(pv_steal_clock, kvm_steal_clock);
 
-		pv_ops.lock.vcpu_is_preempted =
+#ifdef CONFIG_PARAVIRT_SPINLOCKS
+		pv_ops_lock.vcpu_is_preempted =
 			PV_CALLEE_SAVE(__kvm_vcpu_is_preempted);
+#endif
 	}
 
 	if (kvm_para_has_feature(KVM_FEATURE_PV_EOI))
@@ -1121,11 +1123,11 @@ void __init kvm_spinlock_init(void)
 	pr_info("PV spinlocks enabled\n");
 
 	__pv_init_lock_hash();
-	pv_ops.lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
-	pv_ops.lock.queued_spin_unlock =
+	pv_ops_lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
+	pv_ops_lock.queued_spin_unlock =
 		PV_CALLEE_SAVE(__pv_queued_spin_unlock);
-	pv_ops.lock.wait = kvm_wait;
-	pv_ops.lock.kick = kvm_kick_cpu;
+	pv_ops_lock.wait = kvm_wait;
+	pv_ops_lock.kick = kvm_kick_cpu;
 
 	/*
 	 * When PV spinlock is enabled which is preferred over
diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c
index 9e1ea99ad9df..95452444868f 100644
--- a/arch/x86/kernel/paravirt-spinlocks.c
+++ b/arch/x86/kernel/paravirt-spinlocks.c
@@ -3,12 +3,22 @@
  * Split spinlock implementation out into its own file, so it can be
  * compiled in a FTRACE-compatible way.
  */
+#include <linux/static_call.h>
 #include <linux/spinlock.h>
 #include <linux/export.h>
 #include <linux/jump_label.h>
 
-#include <asm/paravirt.h>
+DEFINE_STATIC_KEY_FALSE(virt_spin_lock_key);
 
+#ifdef CONFIG_SMP
+void __init native_pv_lock_init(void)
+{
+	if (boot_cpu_has(X86_FEATURE_HYPERVISOR))
+		static_branch_enable(&virt_spin_lock_key);
+}
+#endif
+
+#ifdef CONFIG_PARAVIRT_SPINLOCKS
 __visible void __native_queued_spin_unlock(struct qspinlock *lock)
 {
 	native_queued_spin_unlock(lock);
@@ -17,7 +27,7 @@ PV_CALLEE_SAVE_REGS_THUNK(__native_queued_spin_unlock);
 
 bool pv_is_native_spin_unlock(void)
 {
-	return pv_ops.lock.queued_spin_unlock.func ==
+	return pv_ops_lock.queued_spin_unlock.func ==
 		__raw_callee_save___native_queued_spin_unlock;
 }
 
@@ -29,7 +39,7 @@ PV_CALLEE_SAVE_REGS_THUNK(__native_vcpu_is_preempted);
 
 bool pv_is_native_vcpu_is_preempted(void)
 {
-	return pv_ops.lock.vcpu_is_preempted.func ==
+	return pv_ops_lock.vcpu_is_preempted.func ==
 		__raw_callee_save___native_vcpu_is_preempted;
 }
 
@@ -41,3 +51,13 @@ void __init paravirt_set_cap(void)
 	if (!pv_is_native_vcpu_is_preempted())
 		setup_force_cpu_cap(X86_FEATURE_VCPUPREEMPT);
 }
+
+struct pv_lock_ops pv_ops_lock = {
+	.queued_spin_lock_slowpath	= native_queued_spin_lock_slowpath,
+	.queued_spin_unlock		= PV_CALLEE_SAVE(__native_queued_spin_unlock),
+	.wait				= paravirt_nop,
+	.kick				= paravirt_nop,
+	.vcpu_is_preempted		= PV_CALLEE_SAVE(__native_vcpu_is_preempted),
+};
+EXPORT_SYMBOL(pv_ops_lock);
+#endif
diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c
index 5dfbd3f55792..a6ed52cae003 100644
--- a/arch/x86/kernel/paravirt.c
+++ b/arch/x86/kernel/paravirt.c
@@ -57,14 +57,6 @@ DEFINE_ASM_FUNC(pv_native_irq_enable, "sti", .noinstr.text);
 DEFINE_ASM_FUNC(pv_native_read_cr2, "mov %cr2, %rax", .noinstr.text);
 #endif
 
-DEFINE_STATIC_KEY_FALSE(virt_spin_lock_key);
-
-void __init native_pv_lock_init(void)
-{
-	if (boot_cpu_has(X86_FEATURE_HYPERVISOR))
-		static_branch_enable(&virt_spin_lock_key);
-}
-
 static noinstr void pv_native_safe_halt(void)
 {
 	native_safe_halt();
@@ -221,19 +213,6 @@ struct paravirt_patch_template pv_ops = {
 
 	.mmu.set_fixmap		= native_set_fixmap,
 #endif /* CONFIG_PARAVIRT_XXL */
-
-#if defined(CONFIG_PARAVIRT_SPINLOCKS)
-	/* Lock ops. */
-#ifdef CONFIG_SMP
-	.lock.queued_spin_lock_slowpath	= native_queued_spin_lock_slowpath,
-	.lock.queued_spin_unlock	=
-				PV_CALLEE_SAVE(__native_queued_spin_unlock),
-	.lock.wait			= paravirt_nop,
-	.lock.kick			= paravirt_nop,
-	.lock.vcpu_is_preempted		=
-				PV_CALLEE_SAVE(__native_vcpu_is_preempted),
-#endif /* SMP */
-#endif
 };
 
 #ifdef CONFIG_PARAVIRT_XXL
diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c
index fe56646d6919..83ac24ead289 100644
--- a/arch/x86/xen/spinlock.c
+++ b/arch/x86/xen/spinlock.c
@@ -134,10 +134,10 @@ void __init xen_init_spinlocks(void)
 	printk(KERN_DEBUG "xen: PV spinlocks enabled\n");
 
 	__pv_init_lock_hash();
-	pv_ops.lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
-	pv_ops.lock.queued_spin_unlock =
+	pv_ops_lock.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath;
+	pv_ops_lock.queued_spin_unlock =
 		PV_CALLEE_SAVE(__pv_queued_spin_unlock);
-	pv_ops.lock.wait = xen_qlock_wait;
-	pv_ops.lock.kick = xen_qlock_kick;
-	pv_ops.lock.vcpu_is_preempted = PV_CALLEE_SAVE(xen_vcpu_stolen);
+	pv_ops_lock.wait = xen_qlock_wait;
+	pv_ops_lock.kick = xen_qlock_kick;
+	pv_ops_lock.vcpu_is_preempted = PV_CALLEE_SAVE(xen_vcpu_stolen);
 }
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index d63d0891924a..36e04988babe 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -550,6 +550,7 @@ static struct {
 	int idx_off;
 } pv_ops_tables[] = {
 	{ .name = "pv_ops", },
+	{ .name = "pv_ops_lock", },
 	{ .name = NULL, .idx_off = -1 }
 };
 
-- 
2.51.0

Re: [PATCH v4 21/21] x86/pvlocks: Move paravirt spinlock functions into own header

[kernel test robot]

Hi Juergen,

kernel test robot noticed the following build errors:

[auto build test ERROR on tip/x86/core]
[also build test ERROR on tip/sched/core kvm/queue kvm/next linus/master v6.18-rc7]
[cannot apply to kvm/linux-next next-20251127]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Juergen-Gross/x86-paravirt-Remove-not-needed-includes-of-paravirt-h/20251127-152054
base:   tip/x86/core
patch link:    https://lore.kernel.org/r/20251127070844.21919-22-jgross%40suse.com
patch subject: [PATCH v4 21/21] x86/pvlocks: Move paravirt spinlock functions into own header
config: x86_64-allnoconfig (https://download.01.org/0day-ci/archive/20251127/202511271747.smpLdjsz-lkp@intel.com/config)
compiler: clang version 20.1.8 (https://github.com/llvm/llvm-project 87f0227cb60147a26a1eeb4fb06e3b505e9c7261)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251127/202511271747.smpLdjsz-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202511271747.smpLdjsz-lkp@intel.com/

All errors (new ones prefixed by >>):

>> arch/x86/kernel/alternative.c:2373:2: error: call to undeclared function 'paravirt_set_cap'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
    2373 |         paravirt_set_cap();
         |         ^
   arch/x86/kernel/alternative.c:2373:2: note: did you mean 'paravirt_ret0'?
   arch/x86/include/asm/paravirt-base.h:23:15: note: 'paravirt_ret0' declared here
      23 | unsigned long paravirt_ret0(void);
         |               ^
   1 error generated.


vim +/paravirt_set_cap +2373 arch/x86/kernel/alternative.c

270a69c4485d7d arch/x86/kernel/alternative.c  Peter Zijlstra            2023-02-08  2344  
9a0b5817ad97bb arch/i386/kernel/alternative.c Gerd Hoffmann             2006-03-23  2345  void __init alternative_instructions(void)
9a0b5817ad97bb arch/i386/kernel/alternative.c Gerd Hoffmann             2006-03-23  2346  {
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2347  	u64 ibt;
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2348  
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2349  	int3_selftest();
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2350  
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2351  	/*
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2352  	 * The patching is not fully atomic, so try to avoid local
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2353  	 * interruptions that might execute the to be patched code.
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2354  	 * Other CPUs are not running.
7457c0da024b18 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2355  	 */
8f4e956b313dcc arch/i386/kernel/alternative.c Andi Kleen                2007-07-22  2356  	stop_nmi();
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2357  
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2358  	/*
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2359  	 * Don't stop machine check exceptions while patching.
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2360  	 * MCEs only happen when something got corrupted and in this
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2361  	 * case we must do something about the corruption.
32b1cbe380417f arch/x86/kernel/alternative.c  Marco Ammon               2019-09-02  2362  	 * Ignoring it is worse than an unlikely patching race.
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2363  	 * Also machine checks tend to be broadcast and if one CPU
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2364  	 * goes into machine check the others follow quickly, so we don't
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2365  	 * expect a machine check to cause undue problems during to code
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2366  	 * patching.
123aa76ec0cab5 arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2367  	 */
8f4e956b313dcc arch/i386/kernel/alternative.c Andi Kleen                2007-07-22  2368  
4e6292114c7412 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2369  	/*
f7af6977621a41 arch/x86/kernel/alternative.c  Juergen Gross             2023-12-10  2370  	 * Make sure to set (artificial) features depending on used paravirt
f7af6977621a41 arch/x86/kernel/alternative.c  Juergen Gross             2023-12-10  2371  	 * functions which can later influence alternative patching.
4e6292114c7412 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2372  	 */
4e6292114c7412 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11 @2373  	paravirt_set_cap();
4e6292114c7412 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2374  
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2375  	/* Keep CET-IBT disabled until caller/callee are patched */
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2376  	ibt = ibt_save(/*disable*/ true);
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2377  
931ab63664f02b arch/x86/kernel/alternative.c  Peter Zijlstra            2022-10-27  2378  	__apply_fineibt(__retpoline_sites, __retpoline_sites_end,
1d7e707af44613 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2379) 			__cfi_sites, __cfi_sites_end, true);
026211c40b0554 arch/x86/kernel/alternative.c  Kees Cook                 2025-09-03  2380  	cfi_debug = false;
931ab63664f02b arch/x86/kernel/alternative.c  Peter Zijlstra            2022-10-27  2381  
7508500900814d arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2382  	/*
7508500900814d arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2383  	 * Rewrite the retpolines, must be done before alternatives since
7508500900814d arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2384  	 * those can rewrite the retpoline thunks.
7508500900814d arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2385  	 */
1d7e707af44613 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2386) 	apply_retpolines(__retpoline_sites, __retpoline_sites_end);
1d7e707af44613 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2387) 	apply_returns(__return_sites, __return_sites_end);
7508500900814d arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2388  
a82b26451de126 arch/x86/kernel/alternative.c  Peter Zijlstra (Intel     2025-06-03  2389) 	its_fini_core();
a82b26451de126 arch/x86/kernel/alternative.c  Peter Zijlstra (Intel     2025-06-03  2390) 
e81dc127ef6988 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2391  	/*
ab9fea59487d8b arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2392  	 * Adjust all CALL instructions to point to func()-10, including
ab9fea59487d8b arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2393  	 * those in .altinstr_replacement.
e81dc127ef6988 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2394  	 */
e81dc127ef6988 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2395  	callthunks_patch_builtin_calls();
e81dc127ef6988 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2396  
ab9fea59487d8b arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2397  	apply_alternatives(__alt_instructions, __alt_instructions_end);
ab9fea59487d8b arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2398  
be0fffa5ca894a arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2399  	/*
be0fffa5ca894a arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2400  	 * Seal all functions that do not have their address taken.
be0fffa5ca894a arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2401  	 */
1d7e707af44613 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2402) 	apply_seal_endbr(__ibt_endbr_seal, __ibt_endbr_seal_end);
ed53a0d971926e arch/x86/kernel/alternative.c  Peter Zijlstra            2022-03-08  2403  
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2404  	ibt_restore(ibt);
ebebe30794d38c arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2405  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: [PATCH v4 21/21] x86/pvlocks: Move paravirt spinlock functions into own header

[kernel test robot]

Hi Juergen,

kernel test robot noticed the following build errors:

[auto build test ERROR on tip/x86/core]
[also build test ERROR on tip/sched/core kvm/queue kvm/next linus/master v6.18-rc7]
[cannot apply to kvm/linux-next next-20251127]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Juergen-Gross/x86-paravirt-Remove-not-needed-includes-of-paravirt-h/20251127-152054
base:   tip/x86/core
patch link:    https://lore.kernel.org/r/20251127070844.21919-22-jgross%40suse.com
patch subject: [PATCH v4 21/21] x86/pvlocks: Move paravirt spinlock functions into own header
config: i386-allnoconfig (https://download.01.org/0day-ci/archive/20251127/202511271704.MdDOB4pB-lkp@intel.com/config)
compiler: gcc-14 (Debian 14.2.0-19) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251127/202511271704.MdDOB4pB-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202511271704.MdDOB4pB-lkp@intel.com/

All errors (new ones prefixed by >>):

   arch/x86/kernel/alternative.c: In function 'alternative_instructions':
>> arch/x86/kernel/alternative.c:2373:9: error: implicit declaration of function 'paravirt_set_cap'; did you mean 'paravirt_ret0'? [-Wimplicit-function-declaration]
    2373 |         paravirt_set_cap();
         |         ^~~~~~~~~~~~~~~~
         |         paravirt_ret0


vim +2373 arch/x86/kernel/alternative.c

270a69c4485d7d0 arch/x86/kernel/alternative.c  Peter Zijlstra            2023-02-08  2344  
9a0b5817ad97bb7 arch/i386/kernel/alternative.c Gerd Hoffmann             2006-03-23  2345  void __init alternative_instructions(void)
9a0b5817ad97bb7 arch/i386/kernel/alternative.c Gerd Hoffmann             2006-03-23  2346  {
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2347  	u64 ibt;
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2348  
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2349  	int3_selftest();
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2350  
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2351  	/*
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2352  	 * The patching is not fully atomic, so try to avoid local
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2353  	 * interruptions that might execute the to be patched code.
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2354  	 * Other CPUs are not running.
7457c0da024b181 arch/x86/kernel/alternative.c  Peter Zijlstra            2019-05-03  2355  	 */
8f4e956b313dccc arch/i386/kernel/alternative.c Andi Kleen                2007-07-22  2356  	stop_nmi();
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2357  
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2358  	/*
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2359  	 * Don't stop machine check exceptions while patching.
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2360  	 * MCEs only happen when something got corrupted and in this
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2361  	 * case we must do something about the corruption.
32b1cbe380417f2 arch/x86/kernel/alternative.c  Marco Ammon               2019-09-02  2362  	 * Ignoring it is worse than an unlikely patching race.
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2363  	 * Also machine checks tend to be broadcast and if one CPU
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2364  	 * goes into machine check the others follow quickly, so we don't
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2365  	 * expect a machine check to cause undue problems during to code
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2366  	 * patching.
123aa76ec0cab5d arch/x86/kernel/alternative.c  Andi Kleen                2009-02-12  2367  	 */
8f4e956b313dccc arch/i386/kernel/alternative.c Andi Kleen                2007-07-22  2368  
4e6292114c74122 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2369  	/*
f7af6977621a416 arch/x86/kernel/alternative.c  Juergen Gross             2023-12-10  2370  	 * Make sure to set (artificial) features depending on used paravirt
f7af6977621a416 arch/x86/kernel/alternative.c  Juergen Gross             2023-12-10  2371  	 * functions which can later influence alternative patching.
4e6292114c74122 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2372  	 */
4e6292114c74122 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11 @2373  	paravirt_set_cap();
4e6292114c74122 arch/x86/kernel/alternative.c  Juergen Gross             2021-03-11  2374  
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2375  	/* Keep CET-IBT disabled until caller/callee are patched */
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2376  	ibt = ibt_save(/*disable*/ true);
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2377  
931ab63664f02b1 arch/x86/kernel/alternative.c  Peter Zijlstra            2022-10-27  2378  	__apply_fineibt(__retpoline_sites, __retpoline_sites_end,
1d7e707af446134 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2379) 			__cfi_sites, __cfi_sites_end, true);
026211c40b05548 arch/x86/kernel/alternative.c  Kees Cook                 2025-09-03  2380  	cfi_debug = false;
931ab63664f02b1 arch/x86/kernel/alternative.c  Peter Zijlstra            2022-10-27  2381  
7508500900814d1 arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2382  	/*
7508500900814d1 arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2383  	 * Rewrite the retpolines, must be done before alternatives since
7508500900814d1 arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2384  	 * those can rewrite the retpoline thunks.
7508500900814d1 arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2385  	 */
1d7e707af446134 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2386) 	apply_retpolines(__retpoline_sites, __retpoline_sites_end);
1d7e707af446134 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2387) 	apply_returns(__return_sites, __return_sites_end);
7508500900814d1 arch/x86/kernel/alternative.c  Peter Zijlstra            2021-10-26  2388  
a82b26451de126a arch/x86/kernel/alternative.c  Peter Zijlstra (Intel     2025-06-03  2389) 	its_fini_core();
a82b26451de126a arch/x86/kernel/alternative.c  Peter Zijlstra (Intel     2025-06-03  2390) 
e81dc127ef69887 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2391  	/*
ab9fea59487d8b5 arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2392  	 * Adjust all CALL instructions to point to func()-10, including
ab9fea59487d8b5 arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2393  	 * those in .altinstr_replacement.
e81dc127ef69887 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2394  	 */
e81dc127ef69887 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2395  	callthunks_patch_builtin_calls();
e81dc127ef69887 arch/x86/kernel/alternative.c  Thomas Gleixner           2022-09-15  2396  
ab9fea59487d8b5 arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2397  	apply_alternatives(__alt_instructions, __alt_instructions_end);
ab9fea59487d8b5 arch/x86/kernel/alternative.c  Peter Zijlstra            2025-02-07  2398  
be0fffa5ca894a9 arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2399  	/*
be0fffa5ca894a9 arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2400  	 * Seal all functions that do not have their address taken.
be0fffa5ca894a9 arch/x86/kernel/alternative.c  Peter Zijlstra            2023-06-22  2401  	 */
1d7e707af446134 arch/x86/kernel/alternative.c  Mike Rapoport (Microsoft  2025-01-26  2402) 	apply_seal_endbr(__ibt_endbr_seal, __ibt_endbr_seal_end);
ed53a0d971926e4 arch/x86/kernel/alternative.c  Peter Zijlstra            2022-03-08  2403  
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2404  	ibt_restore(ibt);
ebebe30794d38c5 arch/x86/kernel/alternative.c  Pawan Gupta               2025-05-03  2405  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: [PATCH v4 00/21] paravirt: cleanup and reorg

[Juergen Gross]

[-- Attachment #1.1.1: Type: text/plain, Size: 7796 bytes --]

At least for patches 1-12: Ping?

Would be nice to have some feedback for 13-21, too (patch 21 needs an
update, though).


Juergen

On 27.11.25 08:08, Juergen Gross wrote:
> Some cleanups and reorg of paravirt code and headers:
> 
> - The first 2 patches should be not controversial at all, as they
>    remove just some no longer needed #include and struct forward
>    declarations.
> 
> - The 3rd patch is removing CONFIG_PARAVIRT_DEBUG, which IMO has
>    no real value, as it just changes a crash to a BUG() (the stack
>    trace will basically be the same). As the maintainer of the main
>    paravirt user (Xen) I have never seen this crash/BUG() to happen.
> 
> - The 4th patch is just a movement of code.
> 
> - I don't know for what reason asm/paravirt_api_clock.h was added,
>    as all archs supporting it do it exactly in the same way. Patch
>    5 is removing it.
> 
> - Patches 6-14 are streamlining the paravirt clock interfaces by
>    using a common implementation across architectures where possible
>    and by moving the related code into common sched code, as this is
>    where it should live.
> 
> - Patches 15-20 are more like RFC material preparing the paravirt
>    infrastructure to support multiple pv_ops function arrays.
>    As a prerequisite for that it makes life in objtool much easier
>    with dropping the Xen static initializers of the pv_ops sub-
>    structures, which is done in patches 15-17.
>    Patches 18-20 are doing the real preparations for multiple pv_ops
>    arrays and using those arrays in multiple headers.
> 
> - Patch 21 is an example how the new scheme can look like using the
>    PV-spinlocks.
> 
> Changes in V2:
> - new patches 13-18 and 20
> - complete rework of patch 21
> 
> Changes in V3:
> - fixed 2 issues detected by kernel test robot
> 
> Changes in V4:
> - fixed one build issue
> 
> Juergen Gross (21):
>    x86/paravirt: Remove not needed includes of paravirt.h
>    x86/paravirt: Remove some unneeded struct declarations
>    x86/paravirt: Remove PARAVIRT_DEBUG config option
>    x86/paravirt: Move thunk macros to paravirt_types.h
>    paravirt: Remove asm/paravirt_api_clock.h
>    sched: Move clock related paravirt code to kernel/sched
>    arm/paravirt: Use common code for paravirt_steal_clock()
>    arm64/paravirt: Use common code for paravirt_steal_clock()
>    loongarch/paravirt: Use common code for paravirt_steal_clock()
>    riscv/paravirt: Use common code for paravirt_steal_clock()
>    x86/paravirt: Use common code for paravirt_steal_clock()
>    x86/paravirt: Move paravirt_sched_clock() related code into tsc.c
>    x86/paravirt: Introduce new paravirt-base.h header
>    x86/paravirt: Move pv_native_*() prototypes to paravirt.c
>    x86/xen: Drop xen_irq_ops
>    x86/xen: Drop xen_cpu_ops
>    x86/xen: Drop xen_mmu_ops
>    objtool: Allow multiple pv_ops arrays
>    x86/paravirt: Allow pv-calls outside paravirt.h
>    x86/paravirt: Specify pv_ops array in paravirt macros
>    x86/pvlocks: Move paravirt spinlock functions into own header
> 
>   arch/Kconfig                                  |   3 +
>   arch/arm/Kconfig                              |   1 +
>   arch/arm/include/asm/paravirt.h               |  22 --
>   arch/arm/include/asm/paravirt_api_clock.h     |   1 -
>   arch/arm/kernel/Makefile                      |   1 -
>   arch/arm/kernel/paravirt.c                    |  23 --
>   arch/arm64/Kconfig                            |   1 +
>   arch/arm64/include/asm/paravirt.h             |  14 -
>   arch/arm64/include/asm/paravirt_api_clock.h   |   1 -
>   arch/arm64/kernel/paravirt.c                  |  11 +-
>   arch/loongarch/Kconfig                        |   1 +
>   arch/loongarch/include/asm/paravirt.h         |  13 -
>   .../include/asm/paravirt_api_clock.h          |   1 -
>   arch/loongarch/kernel/paravirt.c              |  10 +-
>   arch/powerpc/include/asm/paravirt.h           |   3 -
>   arch/powerpc/include/asm/paravirt_api_clock.h |   2 -
>   arch/powerpc/platforms/pseries/setup.c        |   4 +-
>   arch/riscv/Kconfig                            |   1 +
>   arch/riscv/include/asm/paravirt.h             |  14 -
>   arch/riscv/include/asm/paravirt_api_clock.h   |   1 -
>   arch/riscv/kernel/paravirt.c                  |  11 +-
>   arch/x86/Kconfig                              |   8 +-
>   arch/x86/entry/entry_64.S                     |   1 -
>   arch/x86/entry/vsyscall/vsyscall_64.c         |   1 -
>   arch/x86/hyperv/hv_spinlock.c                 |  11 +-
>   arch/x86/include/asm/apic.h                   |   4 -
>   arch/x86/include/asm/highmem.h                |   1 -
>   arch/x86/include/asm/mshyperv.h               |   1 -
>   arch/x86/include/asm/paravirt-base.h          |  29 ++
>   arch/x86/include/asm/paravirt-spinlock.h      | 146 ++++++++
>   arch/x86/include/asm/paravirt.h               | 331 +++++-------------
>   arch/x86/include/asm/paravirt_api_clock.h     |   1 -
>   arch/x86/include/asm/paravirt_types.h         | 269 +++++++-------
>   arch/x86/include/asm/pgtable_32.h             |   1 -
>   arch/x86/include/asm/ptrace.h                 |   2 +-
>   arch/x86/include/asm/qspinlock.h              |  89 +----
>   arch/x86/include/asm/spinlock.h               |   1 -
>   arch/x86/include/asm/timer.h                  |   1 +
>   arch/x86/include/asm/tlbflush.h               |   4 -
>   arch/x86/kernel/Makefile                      |   2 +-
>   arch/x86/kernel/apm_32.c                      |   1 -
>   arch/x86/kernel/callthunks.c                  |   1 -
>   arch/x86/kernel/cpu/bugs.c                    |   1 -
>   arch/x86/kernel/cpu/vmware.c                  |   1 +
>   arch/x86/kernel/kvm.c                         |  13 +-
>   arch/x86/kernel/kvmclock.c                    |   1 +
>   arch/x86/kernel/paravirt-spinlocks.c          |  26 +-
>   arch/x86/kernel/paravirt.c                    |  42 +--
>   arch/x86/kernel/tsc.c                         |  10 +-
>   arch/x86/kernel/vsmp_64.c                     |   1 -
>   arch/x86/lib/cache-smp.c                      |   1 -
>   arch/x86/mm/init.c                            |   1 -
>   arch/x86/xen/enlighten_pv.c                   |  82 ++---
>   arch/x86/xen/irq.c                            |  20 +-
>   arch/x86/xen/mmu_pv.c                         | 100 ++----
>   arch/x86/xen/spinlock.c                       |  11 +-
>   arch/x86/xen/time.c                           |   2 +
>   drivers/clocksource/hyperv_timer.c            |   2 +
>   drivers/xen/time.c                            |   2 +-
>   include/linux/sched/cputime.h                 |  18 +
>   kernel/sched/core.c                           |   5 +
>   kernel/sched/cputime.c                        |  13 +
>   kernel/sched/sched.h                          |   3 +-
>   tools/objtool/arch/x86/decode.c               |   8 +-
>   tools/objtool/check.c                         |  78 ++++-
>   tools/objtool/include/objtool/check.h         |   2 +
>   66 files changed, 661 insertions(+), 826 deletions(-)
>   delete mode 100644 arch/arm/include/asm/paravirt.h
>   delete mode 100644 arch/arm/include/asm/paravirt_api_clock.h
>   delete mode 100644 arch/arm/kernel/paravirt.c
>   delete mode 100644 arch/arm64/include/asm/paravirt_api_clock.h
>   delete mode 100644 arch/loongarch/include/asm/paravirt_api_clock.h
>   delete mode 100644 arch/powerpc/include/asm/paravirt_api_clock.h
>   delete mode 100644 arch/riscv/include/asm/paravirt_api_clock.h
>   create mode 100644 arch/x86/include/asm/paravirt-base.h
>   create mode 100644 arch/x86/include/asm/paravirt-spinlock.h
>   delete mode 100644 arch/x86/include/asm/paravirt_api_clock.h
> 


[-- Attachment #1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 3743 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 495 bytes --]