s390: convert to GENERIC_VDSO

Convert s390 to generic vDSO. There are a few special things on s390:

- vDSO can be called without a stack frame - glibc did this in the past.
  So we need to allocate a stackframe on our own.

- The former assembly code used stcke to get the TOD clock and applied
  time steering to it. We need to do the same in the new code. This is done
  in the architecture specific __arch_get_hw_counter function. The steering
  information is stored in an architecure specific area in the vDSO data.

- CPUCLOCK_VIRT is now handled with a syscall fallback, which might
  be slower/less accurate than the old implementation.

The getcpu() function stays as an assembly function because there is no
generic implementation and the code is just a few lines.

Performance number from my system do 100 mio gettimeofday() calls:

Plain syscall: 8.6s
Generic VDSO:  1.3s
old ASM VDSO:  1s

So it's a bit slower but still much faster than syscalls.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>

6 files changed, 71 insertions, 288 deletions

diff --git a/arch/s390/kernel/vdso64/Makefile b/arch/s390/kernel/vdso64/Makefile
index 4a66a1cb919b..f1c0570780d1 100644
--- a/arch/s390/kernel/vdso64/Makefile
+++ b/arch/s390/kernel/vdso64/Makefile
@@ -1,14 +1,20 @@
 # SPDX-License-Identifier: GPL-2.0
-# List of files in the vdso, has to be asm only for now
+# List of files in the vdso
 
 KCOV_INSTRUMENT := n
+ARCH_REL_TYPE_ABS := R_390_COPY|R_390_GLOB_DAT|R_390_JMP_SLOT|R_390_RELATIVE
+ARCH_REL_TYPE_ABS += R_390_GOT|R_390_PLT
 
-obj-vdso64 = gettimeofday.o clock_getres.o clock_gettime.o note.o getcpu.o
+include $(srctree)/lib/vdso/Makefile
+obj-vdso64 = vdso_user_wrapper.o note.o getcpu.o
+obj-cvdso64 = vdso64_generic.o
+CFLAGS_REMOVE_vdso64_generic.o = -pg $(CC_FLAGS_FTRACE) $(CC_FLAGS_EXPOLINE)
 
 # Build rules
 
-targets := $(obj-vdso64) vdso64.so vdso64.so.dbg
+targets := $(obj-vdso64) $(obj-cvdso64) vdso64.so vdso64.so.dbg
 obj-vdso64 := $(addprefix $(obj)/, $(obj-vdso64))
+obj-cvdso64 := $(addprefix $(obj)/, $(obj-cvdso64))
 
 KBUILD_AFLAGS += -DBUILD_VDSO
 KBUILD_CFLAGS += -DBUILD_VDSO
@@ -37,7 +43,7 @@ KASAN_SANITIZE := n
 $(obj)/vdso64_wrapper.o : $(obj)/vdso64.so
 
 # link rule for the .so file, .lds has to be first
-$(obj)/vdso64.so.dbg: $(obj)/vdso64.lds $(obj-vdso64) FORCE
+$(obj)/vdso64.so.dbg: $(src)/vdso64.lds $(obj-vdso64) $(obj-cvdso64) FORCE
 	$(call if_changed,ld)
 
 # strip rule for the .so file
@@ -49,9 +55,14 @@ $(obj)/%.so: $(obj)/%.so.dbg FORCE
 $(obj-vdso64): %.o: %.S FORCE
 	$(call if_changed_dep,vdso64as)
 
+$(obj-cvdso64): %.o: %.c FORCE
+	$(call if_changed_dep,vdso64cc)
+
 # actual build commands
 quiet_cmd_vdso64as = VDSO64A $@
       cmd_vdso64as = $(CC) $(a_flags) -c -o $@ $<
+quiet_cmd_vdso64cc = VDSO64C $@
+      cmd_vdso64cc = $(CC) $(c_flags) -c -o $@ $<
 
 # install commands for the unstripped file
 quiet_cmd_vdso_install = INSTALL $@
diff --git a/arch/s390/kernel/vdso64/clock_getres.S b/arch/s390/kernel/vdso64/clock_getres.S
deleted file mode 100644
index 0c79caa32b59..000000000000
--- a/arch/s390/kernel/vdso64/clock_getres.S
+++ /dev/null
@@ -1,50 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * Userland implementation of clock_getres() for 64 bits processes in a
- * s390 kernel for use in the vDSO
- *
- *  Copyright IBM Corp. 2008
- *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
- */
-#include <asm/vdso.h>
-#include <asm/asm-offsets.h>
-#include <asm/unistd.h>
-#include <asm/dwarf.h>
-
-	.text
-	.align 4
-	.globl __kernel_clock_getres
-	.type  __kernel_clock_getres,@function
-__kernel_clock_getres:
-	CFI_STARTPROC
-	larl	%r1,3f
-	lg	%r0,0(%r1)
-	cghi	%r2,__CLOCK_REALTIME_COARSE
-	je	0f
-	cghi	%r2,__CLOCK_MONOTONIC_COARSE
-	je	0f
-	larl	%r1,_vdso_data
-	llgf	%r0,__VDSO_CLOCK_REALTIME_RES(%r1)
-	cghi	%r2,__CLOCK_REALTIME
-	je	0f
-	cghi	%r2,__CLOCK_MONOTONIC
-	je	0f
-	cghi	%r2,__CLOCK_THREAD_CPUTIME_ID
-	je	0f
-	cghi	%r2,-2		/* Per-thread CPUCLOCK with PID=0, VIRT=1 */
-	jne	2f
-	larl	%r5,_vdso_data
-	icm	%r0,15,__LC_ECTG_OK(%r5)
-	jz	2f
-0:	ltgr	%r3,%r3
-	jz	1f				/* res == NULL */
-	xc	0(8,%r3),0(%r3)			/* set tp->tv_sec to zero */
-	stg	%r0,8(%r3)			/* store tp->tv_usec */
-1:	lghi	%r2,0
-	br	%r14
-2:	lghi	%r1,__NR_clock_getres		/* fallback to svc */
-	svc	0
-	br	%r14
-	CFI_ENDPROC
-3:	.quad	__CLOCK_COARSE_RES
-	.size	__kernel_clock_getres,.-__kernel_clock_getres
diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S
deleted file mode 100644
index 9d2ee79b90f2..000000000000
--- a/arch/s390/kernel/vdso64/clock_gettime.S
+++ /dev/null
@@ -1,163 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * Userland implementation of clock_gettime() for 64 bits processes in a
- * s390 kernel for use in the vDSO
- *
- *  Copyright IBM Corp. 2008
- *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
- */
-#include <asm/vdso.h>
-#include <asm/asm-offsets.h>
-#include <asm/unistd.h>
-#include <asm/dwarf.h>
-#include <asm/ptrace.h>
-
-	.text
-	.align 4
-	.globl __kernel_clock_gettime
-	.type  __kernel_clock_gettime,@function
-__kernel_clock_gettime:
-	CFI_STARTPROC
-	aghi	%r15,-16
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD+16
-	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
-	larl	%r5,_vdso_data
-	cghi	%r2,__CLOCK_REALTIME_COARSE
-	je	4f
-	cghi	%r2,__CLOCK_REALTIME
-	je	5f
-	cghi	%r2,-3		/* Per-thread CPUCLOCK with PID=0, VIRT=1 */
-	je	9f
-	cghi	%r2,__CLOCK_MONOTONIC_COARSE
-	je	3f
-	cghi	%r2,__CLOCK_MONOTONIC
-	jne	12f
-
-	/* CLOCK_MONOTONIC */
-0:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
-	tmll	%r4,0x0001			/* pending update ? loop */
-	jnz	0b
-	stcke	0(%r15)				/* Store TOD clock */
-	lgf	%r2,__VDSO_TK_SHIFT(%r5)	/* Timekeeper shift */
-	lg	%r0,__VDSO_WTOM_SEC(%r5)
-	lg	%r1,1(%r15)
-	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
-	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
-	alg	%r1,__VDSO_WTOM_NSEC(%r5)
-	srlg	%r1,%r1,0(%r2)			/*  >> tk->shift */
-	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
-	jne	0b
-	larl	%r5,13f
-1:	clg	%r1,0(%r5)
-	jl	2f
-	slg	%r1,0(%r5)
-	aghi	%r0,1
-	j	1b
-2:	stg	%r0,0(%r3)			/* store tp->tv_sec */
-	stg	%r1,8(%r3)			/* store tp->tv_nsec */
-	lghi	%r2,0
-	aghi	%r15,16
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD
-	CFI_RESTORE 15
-	br	%r14
-
-	/* CLOCK_MONOTONIC_COARSE */
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD+16
-	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
-3:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
-	tmll	%r4,0x0001			/* pending update ? loop */
-	jnz	3b
-	lg	%r0,__VDSO_WTOM_CRS_SEC(%r5)
-	lg	%r1,__VDSO_WTOM_CRS_NSEC(%r5)
-	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
-	jne	3b
-	j	2b
-
-	/* CLOCK_REALTIME_COARSE */
-4:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
-	tmll	%r4,0x0001			/* pending update ? loop */
-	jnz	4b
-	lg	%r0,__VDSO_XTIME_CRS_SEC(%r5)
-	lg	%r1,__VDSO_XTIME_CRS_NSEC(%r5)
-	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
-	jne	4b
-	j	7f
-
-	/* CLOCK_REALTIME */
-5:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
-	tmll	%r4,0x0001			/* pending update ? loop */
-	jnz	5b
-	stcke	0(%r15)				/* Store TOD clock */
-	lg	%r1,1(%r15)
-	lg	%r0,__VDSO_TS_END(%r5)		/* TOD steering end time */
-	slgr	%r0,%r1				/* now - ts_steering_end */
-	ltgr	%r0,%r0				/* past end of steering ? */
-	jm	17f
-	srlg	%r0,%r0,15			/* 1 per 2^16 */
-	tm	__VDSO_TS_DIR+3(%r5),0x01	/* steering direction? */
-	jz	18f
-	lcgr	%r0,%r0				/* negative TOD offset */
-18:	algr	%r1,%r0				/* add steering offset */
-17:	lgf	%r2,__VDSO_TK_SHIFT(%r5)	/* Timekeeper shift */
-	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
-	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
-	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + tk->xtime_nsec */
-	srlg	%r1,%r1,0(%r2)			/*  >> tk->shift */
-	lg	%r0,__VDSO_XTIME_SEC(%r5)	/* tk->xtime_sec */
-	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
-	jne	5b
-	larl	%r5,13f
-6:	clg	%r1,0(%r5)
-	jl	7f
-	slg	%r1,0(%r5)
-	aghi	%r0,1
-	j	6b
-7:	stg	%r0,0(%r3)			/* store tp->tv_sec */
-	stg	%r1,8(%r3)			/* store tp->tv_nsec */
-	lghi	%r2,0
-	aghi	%r15,16
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD
-	CFI_RESTORE 15
-	br	%r14
-
-	/* CPUCLOCK_VIRT for this thread */
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD+16
-	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
-9:	lghi	%r4,0
-	icm	%r0,15,__VDSO_ECTG_OK(%r5)
-	jz	12f
-	sacf	256				/* Magic ectg instruction */
-	.insn	ssf,0xc80100000000,__VDSO_ECTG_BASE(4),__VDSO_ECTG_USER(4),4
-	sacf	0
-	algr	%r1,%r0				/* r1 = cputime as TOD value */
-	mghi	%r1,1000			/* convert to nanoseconds */
-	srlg	%r1,%r1,12			/* r1 = cputime in nanosec */
-	lgr	%r4,%r1
-	larl	%r5,13f
-	srlg	%r1,%r1,9			/* divide by 1000000000 */
-	mlg	%r0,8(%r5)
-	srlg	%r0,%r0,11			/* r0 = tv_sec */
-	stg	%r0,0(%r3)
-	msg	%r0,0(%r5)			/* calculate tv_nsec */
-	slgr	%r4,%r0				/* r4 = tv_nsec */
-	stg	%r4,8(%r3)
-	lghi	%r2,0
-	aghi	%r15,16
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD
-	CFI_RESTORE 15
-	br	%r14
-
-	/* Fallback to system call */
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD+16
-	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
-12:	lghi	%r1,__NR_clock_gettime
-	svc	0
-	aghi	%r15,16
-	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD
-	CFI_RESTORE 15
-	br	%r14
-	CFI_ENDPROC
-
-13:	.quad	1000000000
-14:	.quad	19342813113834067
-	.size	__kernel_clock_gettime,.-__kernel_clock_gettime
diff --git a/arch/s390/kernel/vdso64/gettimeofday.S b/arch/s390/kernel/vdso64/gettimeofday.S
deleted file mode 100644
index aebe10dc7c99..000000000000
--- a/arch/s390/kernel/vdso64/gettimeofday.S
+++ /dev/null
@@ -1,71 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * Userland implementation of gettimeofday() for 64 bits processes in a
- * s390 kernel for use in the vDSO
- *
- *  Copyright IBM Corp. 2008
- *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
- */
-#include <asm/vdso.h>
-#include <asm/asm-offsets.h>
-#include <asm/unistd.h>
-#include <asm/dwarf.h>
-#include <asm/ptrace.h>
-
-	.text
-	.align 4
-	.globl __kernel_gettimeofday
-	.type  __kernel_gettimeofday,@function
-__kernel_gettimeofday:
-	CFI_STARTPROC
-	aghi	%r15,-16
-	CFI_ADJUST_CFA_OFFSET 16
-	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
-	larl	%r5,_vdso_data
-0:	ltgr	%r3,%r3				/* check if tz is NULL */
-	je	1f
-	mvc	0(8,%r3),__VDSO_TIMEZONE(%r5)
-1:	ltgr	%r2,%r2				/* check if tv is NULL */
-	je	4f
-	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
-	tmll	%r4,0x0001			/* pending update ? loop */
-	jnz	0b
-	stcke	0(%r15)				/* Store TOD clock */
-	lg	%r1,1(%r15)
-	lg	%r0,__VDSO_TS_END(%r5)		/* TOD steering end time */
-	slgr	%r0,%r1				/* now - ts_steering_end */
-	ltgr	%r0,%r0				/* past end of steering ? */
-	jm	6f
-	srlg	%r0,%r0,15			/* 1 per 2^16 */
-	tm	__VDSO_TS_DIR+3(%r5),0x01	/* steering direction? */
-	jz	7f
-	lcgr	%r0,%r0				/* negative TOD offset */
-7:	algr	%r1,%r0				/* add steering offset */
-6:	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
-	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
-	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + tk->xtime_nsec */
-	lg	%r0,__VDSO_XTIME_SEC(%r5)	/* tk->xtime_sec */
-	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
-	jne	0b
-	lgf	%r5,__VDSO_TK_SHIFT(%r5)	/* Timekeeper shift */
-	srlg	%r1,%r1,0(%r5)			/*  >> tk->shift */
-	larl	%r5,5f
-2:	clg	%r1,0(%r5)
-	jl	3f
-	slg	%r1,0(%r5)
-	aghi	%r0,1
-	j	2b
-3:	stg	%r0,0(%r2)			/* store tv->tv_sec */
-	slgr	%r0,%r0				/* tv_nsec -> tv_usec */
-	ml	%r0,8(%r5)
-	srlg	%r0,%r0,6
-	stg	%r0,8(%r2)			/* store tv->tv_usec */
-4:	lghi	%r2,0
-	aghi	%r15,16
-	CFI_ADJUST_CFA_OFFSET -16
-	CFI_RESTORE 15
-	br	%r14
-	CFI_ENDPROC
-5:	.quad	1000000000
-	.long	274877907
-	.size	__kernel_gettimeofday,.-__kernel_gettimeofday
diff --git a/arch/s390/kernel/vdso64/vdso64_generic.c b/arch/s390/kernel/vdso64/vdso64_generic.c
new file mode 100644
index 000000000000..a8cef7e4d137
--- /dev/null
+++ b/arch/s390/kernel/vdso64/vdso64_generic.c
@@ -0,0 +1,18 @@
+// SPDX-License-Identifier: GPL-2.0
+#include "../../../../lib/vdso/gettimeofday.c"
+
+int __s390_vdso_gettimeofday(struct __kernel_old_timeval *tv,
+			     struct timezone *tz)
+{
+	return __cvdso_gettimeofday(tv, tz);
+}
+
+int __s390_vdso_clock_gettime(clockid_t clock, struct __kernel_timespec *ts)
+{
+	return __cvdso_clock_gettime(clock, ts);
+}
+
+int __s390_vdso_clock_getres(clockid_t clock, struct __kernel_timespec *ts)
+{
+	return __cvdso_clock_getres(clock, ts);
+}
diff --git a/arch/s390/kernel/vdso64/vdso_user_wrapper.S b/arch/s390/kernel/vdso64/vdso_user_wrapper.S
new file mode 100644
index 000000000000..a775d7e52872
--- /dev/null
+++ b/arch/s390/kernel/vdso64/vdso_user_wrapper.S
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#include <asm/vdso.h>
+#include <asm/unistd.h>
+#include <asm/asm-offsets.h>
+#include <asm/dwarf.h>
+#include <asm/ptrace.h>
+
+#define WRAPPER_FRAME_SIZE (STACK_FRAME_OVERHEAD+8)
+
+/*
+ * Older glibc version called vdso without allocating a stackframe. This wrapper
+ * is just used to allocate a stackframe. See
+ * https://sourceware.org/git/?p=glibc.git;a=commit;h=478593e6374f3818da39332260dc453cb19cfa1e
+ * for details.
+ */
+.macro vdso_func func
+	.globl __kernel_\func
+	.type  __kernel_\func,@function
+	.align 8
+__kernel_\func:
+	CFI_STARTPROC
+	aghi	%r15,-WRAPPER_FRAME_SIZE
+	CFI_DEF_CFA_OFFSET (STACK_FRAME_OVERHEAD + WRAPPER_FRAME_SIZE)
+	CFI_VAL_OFFSET 15, -STACK_FRAME_OVERHEAD
+	stg	%r14,STACK_FRAME_OVERHEAD(%r15)
+	brasl	%r14,__s390_vdso_\func
+	lg	%r14,STACK_FRAME_OVERHEAD(%r15)
+	aghi	%r15,WRAPPER_FRAME_SIZE
+	CFI_DEF_CFA_OFFSET STACK_FRAME_OVERHEAD
+	CFI_RESTORE 15
+	br	%r14
+	CFI_ENDPROC
+	.size	__kernel_\func,.-__kernel_\func
+.endm
+
+vdso_func gettimeofday
+vdso_func clock_getres
+vdso_func clock_gettime