summaryrefslogtreecommitdiff
path: root/arch/sparc/vdso/vclock_gettime.c
diff options
context:
space:
mode:
authorNagarathnam Muthusamy <nagarathnam.muthusamy@oracle.com>2017-09-21 11:05:31 -0400
committerDavid S. Miller <davem@davemloft.net>2017-11-15 14:21:03 +0900
commit9a08862a5d2e266ecea1865547463da2745fc687 (patch)
treec651ef6b7c2d7dc75e0ec73fcc6d2e2f15f72a9e /arch/sparc/vdso/vclock_gettime.c
parent23198ddffb6cddb5d5824230af4dd4b46e4046a4 (diff)
vDSO for sparc
Following patch is based on work done by Nick Alcock on 64-bit vDSO for sparc in Oracle linux. I have extended it to include support for 32-bit vDSO for sparc on 64-bit kernel. vDSO for sparc is based on the X86 implementation. This patch provides vDSO support for both 64-bit and 32-bit programs on 64-bit kernel. vDSO will be disabled on 32-bit linux kernel on sparc. *) vclock_gettime.c contains all the vdso functions. Since data page is mapped before the vdso code page, the pointer to data page is got by subracting offset from an address in the vdso code page. The return address stored in %i7 is used for this purpose. *) During compilation, both 32-bit and 64-bit vdso images are compiled and are converted into raw bytes by vdso2c program to be ready for mapping into the process. 32-bit images are compiled only if CONFIG_COMPAT is enabled. vdso2c generates two files vdso-image-64.c and vdso-image-32.c which contains the respective vDSO image in C structure. *) During vdso initialization, required number of vdso pages are allocated and raw bytes are copied into the pages. *) During every exec, these pages are mapped into the process through arch_setup_additional_pages and the location of mapping is passed on to the process through aux vector AT_SYSINFO_EHDR which is used by glibc. *) A new update_vsyscall routine for sparc is added to keep the data page in vdso updated. *) As vDSO cannot contain dynamically relocatable references, a new version of cpu_relax is added for the use of vDSO. This change also requires a putback to glibc to use vDSO. For testing, programs planning to try vDSO can be compiled against the generated vdso(64/32).so in the source. Testing: ======== [root@localhost ~]# cat vdso_test.c int main() { struct timespec tv_start, tv_end; struct timeval tv_tmp; int i; int count = 1 * 1000 * 10000; long long diff; clock_gettime(0, &tv_start); for (i = 0; i < count; i++) gettimeofday(&tv_tmp, NULL); clock_gettime(0, &tv_end); diff = (long long)(tv_end.tv_sec - tv_start.tv_sec)*(1*1000*1000*1000); diff += (tv_end.tv_nsec - tv_start.tv_nsec); printf("Start sec: %d\n", tv_start.tv_sec); printf("End sec : %d\n", tv_end.tv_sec); printf("%d cycles in %lld ns = %f ns/cycle\n", count, diff, (double)diff / (double)count); return 0; } [root@localhost ~]# cc vdso_test.c -o t32_without_fix -m32 -lrt [root@localhost ~]# ./t32_without_fix Start sec: 1502396130 End sec : 1502396140 10000000 cycles in 9565148528 ns = 956.514853 ns/cycle [root@localhost ~]# cc vdso_test.c -o t32_with_fix -m32 ./vdso32.so.dbg [root@localhost ~]# ./t32_with_fix Start sec: 1502396168 End sec : 1502396169 10000000 cycles in 798141262 ns = 79.814126 ns/cycle [root@localhost ~]# cc vdso_test.c -o t64_without_fix -m64 -lrt [root@localhost ~]# ./t64_without_fix Start sec: 1502396208 End sec : 1502396218 10000000 cycles in 9846091800 ns = 984.609180 ns/cycle [root@localhost ~]# cc vdso_test.c -o t64_with_fix -m64 ./vdso64.so.dbg [root@localhost ~]# ./t64_with_fix Start sec: 1502396257 End sec : 1502396257 10000000 cycles in 380984048 ns = 38.098405 ns/cycle V1 to V2 Changes: ================= Added hot patching code to switch the read stick instruction to read tick instruction based on the hardware. V2 to V3 Changes: ================= Merged latest changes from sparc-next and moved the initialization of clocksource_tick.archdata.vclock_mode to time_init_early. Disabled queued spinlock and rwlock configuration when simulating 32-bit config to compile 32-bit VDSO. V3 to V4 Changes: ================= Hardcoded the page size as 8192 in linker script for both 64-bit and 32-bit binaries. Removed unused variables in vdso2c.h. Added -mv8plus flag to Makefile to prevent the generation of relocation entries for __lshrdi3 in 32-bit vdso binary. Signed-off-by: Nick Alcock <nick.alcock@oracle.com> Signed-off-by: Nagarathnam Muthusamy <nagarathnam.muthusamy@oracle.com> Reviewed-by: Shannon Nelson <shannon.nelson@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'arch/sparc/vdso/vclock_gettime.c')
-rw-r--r--arch/sparc/vdso/vclock_gettime.c264
1 files changed, 264 insertions, 0 deletions
diff --git a/arch/sparc/vdso/vclock_gettime.c b/arch/sparc/vdso/vclock_gettime.c
new file mode 100644
index 000000000000..3feb3d960ca5
--- /dev/null
+++ b/arch/sparc/vdso/vclock_gettime.c
@@ -0,0 +1,264 @@
+/*
+ * Copyright 2006 Andi Kleen, SUSE Labs.
+ * Subject to the GNU Public License, v.2
+ *
+ * Fast user context implementation of clock_gettime, gettimeofday, and time.
+ *
+ * The code should have no internal unresolved relocations.
+ * Check with readelf after changing.
+ * Also alternative() doesn't work.
+ */
+/*
+ * Copyright (c) 2017 Oracle and/or its affiliates. All rights reserved.
+ */
+
+/* Disable profiling for userspace code: */
+#ifndef DISABLE_BRANCH_PROFILING
+#define DISABLE_BRANCH_PROFILING
+#endif
+
+#include <linux/kernel.h>
+#include <linux/time.h>
+#include <linux/string.h>
+#include <asm/io.h>
+#include <asm/unistd.h>
+#include <asm/timex.h>
+#include <asm/clocksource.h>
+#include <asm/vvar.h>
+
+#undef TICK_PRIV_BIT
+#ifdef CONFIG_SPARC64
+#define TICK_PRIV_BIT (1UL << 63)
+#else
+#define TICK_PRIV_BIT (1ULL << 63)
+#endif
+
+#define SYSCALL_STRING \
+ "ta 0x6d;" \
+ "sub %%g0, %%o0, %%o0;" \
+
+#define SYSCALL_CLOBBERS \
+ "f0", "f1", "f2", "f3", "f4", "f5", "f6", "f7", \
+ "f8", "f9", "f10", "f11", "f12", "f13", "f14", "f15", \
+ "f16", "f17", "f18", "f19", "f20", "f21", "f22", "f23", \
+ "f24", "f25", "f26", "f27", "f28", "f29", "f30", "f31", \
+ "f32", "f34", "f36", "f38", "f40", "f42", "f44", "f46", \
+ "f48", "f50", "f52", "f54", "f56", "f58", "f60", "f62", \
+ "cc", "memory"
+
+/*
+ * Compute the vvar page's address in the process address space, and return it
+ * as a pointer to the vvar_data.
+ */
+static notrace noinline struct vvar_data *
+get_vvar_data(void)
+{
+ unsigned long ret;
+
+ /*
+ * vdso data page is the first vDSO page so grab the return address
+ * and move up a page to get to the data page.
+ */
+ ret = (unsigned long)__builtin_return_address(0);
+ ret &= ~(8192 - 1);
+ ret -= 8192;
+
+ return (struct vvar_data *) ret;
+}
+
+static notrace long
+vdso_fallback_gettime(long clock, struct timespec *ts)
+{
+ register long num __asm__("g1") = __NR_clock_gettime;
+ register long o0 __asm__("o0") = clock;
+ register long o1 __asm__("o1") = (long) ts;
+
+ __asm__ __volatile__(SYSCALL_STRING : "=r" (o0) : "r" (num),
+ "0" (o0), "r" (o1) : SYSCALL_CLOBBERS);
+ return o0;
+}
+
+static notrace __always_inline long
+vdso_fallback_gettimeofday(struct timeval *tv, struct timezone *tz)
+{
+ register long num __asm__("g1") = __NR_gettimeofday;
+ register long o0 __asm__("o0") = (long) tv;
+ register long o1 __asm__("o1") = (long) tz;
+
+ __asm__ __volatile__(SYSCALL_STRING : "=r" (o0) : "r" (num),
+ "0" (o0), "r" (o1) : SYSCALL_CLOBBERS);
+ return o0;
+}
+
+#ifdef CONFIG_SPARC64
+static notrace noinline u64
+vread_tick(void) {
+ u64 ret;
+
+ __asm__ __volatile__("rd %%asr24, %0 \n"
+ ".section .vread_tick_patch, \"ax\" \n"
+ "rd %%tick, %0 \n"
+ ".previous \n"
+ : "=&r" (ret));
+ return ret & ~TICK_PRIV_BIT;
+}
+#else
+static notrace noinline u64
+vread_tick(void)
+{
+ unsigned int lo, hi;
+
+ __asm__ __volatile__("rd %%asr24, %%g1\n\t"
+ "srlx %%g1, 32, %1\n\t"
+ "srl %%g1, 0, %0\n"
+ ".section .vread_tick_patch, \"ax\" \n"
+ "rd %%tick, %%g1\n"
+ ".previous \n"
+ : "=&r" (lo), "=&r" (hi)
+ :
+ : "g1");
+ return lo | ((u64)hi << 32);
+}
+#endif
+
+static notrace inline u64
+vgetsns(struct vvar_data *vvar)
+{
+ u64 v;
+ u64 cycles;
+
+ cycles = vread_tick();
+ v = (cycles - vvar->clock.cycle_last) & vvar->clock.mask;
+ return v * vvar->clock.mult;
+}
+
+static notrace noinline int
+do_realtime(struct vvar_data *vvar, struct timespec *ts)
+{
+ unsigned long seq;
+ u64 ns;
+
+ ts->tv_nsec = 0;
+ do {
+ seq = vvar_read_begin(vvar);
+ ts->tv_sec = vvar->wall_time_sec;
+ ns = vvar->wall_time_snsec;
+ ns += vgetsns(vvar);
+ ns >>= vvar->clock.shift;
+ } while (unlikely(vvar_read_retry(vvar, seq)));
+
+ timespec_add_ns(ts, ns);
+
+ return 0;
+}
+
+static notrace noinline int
+do_monotonic(struct vvar_data *vvar, struct timespec *ts)
+{
+ unsigned long seq;
+ u64 ns;
+
+ ts->tv_nsec = 0;
+ do {
+ seq = vvar_read_begin(vvar);
+ ts->tv_sec = vvar->monotonic_time_sec;
+ ns = vvar->monotonic_time_snsec;
+ ns += vgetsns(vvar);
+ ns >>= vvar->clock.shift;
+ } while (unlikely(vvar_read_retry(vvar, seq)));
+
+ timespec_add_ns(ts, ns);
+
+ return 0;
+}
+
+static notrace noinline int
+do_realtime_coarse(struct vvar_data *vvar, struct timespec *ts)
+{
+ unsigned long seq;
+
+ do {
+ seq = vvar_read_begin(vvar);
+ ts->tv_sec = vvar->wall_time_coarse_sec;
+ ts->tv_nsec = vvar->wall_time_coarse_nsec;
+ } while (unlikely(vvar_read_retry(vvar, seq)));
+ return 0;
+}
+
+static notrace noinline int
+do_monotonic_coarse(struct vvar_data *vvar, struct timespec *ts)
+{
+ unsigned long seq;
+
+ do {
+ seq = vvar_read_begin(vvar);
+ ts->tv_sec = vvar->monotonic_time_coarse_sec;
+ ts->tv_nsec = vvar->monotonic_time_coarse_nsec;
+ } while (unlikely(vvar_read_retry(vvar, seq)));
+
+ return 0;
+}
+
+notrace int
+__vdso_clock_gettime(clockid_t clock, struct timespec *ts)
+{
+ struct vvar_data *vvd = get_vvar_data();
+
+ switch (clock) {
+ case CLOCK_REALTIME:
+ if (unlikely(vvd->vclock_mode == VCLOCK_NONE))
+ break;
+ return do_realtime(vvd, ts);
+ case CLOCK_MONOTONIC:
+ if (unlikely(vvd->vclock_mode == VCLOCK_NONE))
+ break;
+ return do_monotonic(vvd, ts);
+ case CLOCK_REALTIME_COARSE:
+ return do_realtime_coarse(vvd, ts);
+ case CLOCK_MONOTONIC_COARSE:
+ return do_monotonic_coarse(vvd, ts);
+ }
+ /*
+ * Unknown clock ID ? Fall back to the syscall.
+ */
+ return vdso_fallback_gettime(clock, ts);
+}
+int
+clock_gettime(clockid_t, struct timespec *)
+ __attribute__((weak, alias("__vdso_clock_gettime")));
+
+notrace int
+__vdso_gettimeofday(struct timeval *tv, struct timezone *tz)
+{
+ struct vvar_data *vvd = get_vvar_data();
+
+ if (likely(vvd->vclock_mode != VCLOCK_NONE)) {
+ if (likely(tv != NULL)) {
+ union tstv_t {
+ struct timespec ts;
+ struct timeval tv;
+ } *tstv = (union tstv_t *) tv;
+ do_realtime(vvd, &tstv->ts);
+ /*
+ * Assign before dividing to ensure that the division is
+ * done in the type of tv_usec, not tv_nsec.
+ *
+ * There cannot be > 1 billion usec in a second:
+ * do_realtime() has already distributed such overflow
+ * into tv_sec. So we can assign it to an int safely.
+ */
+ tstv->tv.tv_usec = tstv->ts.tv_nsec;
+ tstv->tv.tv_usec /= 1000;
+ }
+ if (unlikely(tz != NULL)) {
+ /* Avoid memcpy. Some old compilers fail to inline it */
+ tz->tz_minuteswest = vvd->tz_minuteswest;
+ tz->tz_dsttime = vvd->tz_dsttime;
+ }
+ return 0;
+ }
+ return vdso_fallback_gettimeofday(tv, tz);
+}
+int
+gettimeofday(struct timeval *, struct timezone *)
+ __attribute__((weak, alias("__vdso_gettimeofday")));