[Midnightbsd-cvs] src [8099] trunk/sys: bring in FreeBSD method to export kernel timekeeping data to usermode using a shared page.
laffer1 at midnightbsd.org
laffer1 at midnightbsd.org
Thu Sep 15 19:01:55 EDT 2016
Revision: 8099
http://svnweb.midnightbsd.org/src/?rev=8099
Author: laffer1
Date: 2016-09-15 19:01:55 -0400 (Thu, 15 Sep 2016)
Log Message:
-----------
bring in FreeBSD method to export kernel timekeeping data to usermode using a shared page.
Modified Paths:
--------------
trunk/sys/kern/imgact_elf.c
trunk/sys/kern/kern_exec.c
trunk/sys/sys/sysent.h
trunk/sys/x86/x86/tsc.c
Modified: trunk/sys/kern/imgact_elf.c
===================================================================
--- trunk/sys/kern/imgact_elf.c 2016-09-15 22:56:13 UTC (rev 8098)
+++ trunk/sys/kern/imgact_elf.c 2016-09-15 23:01:55 UTC (rev 8099)
@@ -1005,6 +1005,10 @@
AUXARGS_ENTRY(pos, AT_PAGESIZES, imgp->pagesizes);
AUXARGS_ENTRY(pos, AT_PAGESIZESLEN, imgp->pagesizeslen);
}
+ if (imgp->sysent->sv_timekeep_base != 0) {
+ AUXARGS_ENTRY(pos, AT_TIMEKEEP,
+ imgp->sysent->sv_timekeep_base);
+ }
AUXARGS_ENTRY(pos, AT_STACKPROT, imgp->sysent->sv_shared_page_obj
!= NULL && imgp->stack_prot != 0 ? imgp->stack_prot :
imgp->sysent->sv_stackprot);
Modified: trunk/sys/kern/kern_exec.c
===================================================================
--- trunk/sys/kern/kern_exec.c 2016-09-15 22:56:13 UTC (rev 8098)
+++ trunk/sys/kern/kern_exec.c 2016-09-15 23:01:55 UTC (rev 8099)
@@ -28,6 +28,7 @@
__FBSDID("$FreeBSD$");
#include "opt_capsicum.h"
+#include "opt_compat.h"
#include "opt_hwpmc_hooks.h"
#include "opt_kdtrace.h"
#include "opt_ktrace.h"
@@ -64,6 +65,7 @@
#include <sys/sysent.h>
#include <sys/shm.h>
#include <sys/sysctl.h>
+#include <sys/vdso.h>
#include <sys/vnode.h>
#include <sys/stat.h>
#ifdef KTRACE
@@ -1517,42 +1519,13 @@
static struct sx shared_page_alloc_sx;
static vm_object_t shared_page_obj;
static int shared_page_free;
+char *shared_page_mapping;
-struct sf_buf *
-shared_page_write_start(int base)
-{
- vm_page_t m;
- struct sf_buf *s;
-
- VM_OBJECT_LOCK(shared_page_obj);
- m = vm_page_grab(shared_page_obj, OFF_TO_IDX(base), VM_ALLOC_RETRY);
- VM_OBJECT_UNLOCK(shared_page_obj);
- s = sf_buf_alloc(m, SFB_DEFAULT);
- return (s);
-}
-
void
-shared_page_write_end(struct sf_buf *sf)
-{
- vm_page_t m;
-
- m = sf_buf_page(sf);
- sf_buf_free(sf);
- VM_OBJECT_LOCK(shared_page_obj);
- vm_page_wakeup(m);
- VM_OBJECT_UNLOCK(shared_page_obj);
-}
-
-void
shared_page_write(int base, int size, const void *data)
{
- struct sf_buf *sf;
- vm_offset_t sk;
- sf = shared_page_write_start(base);
- sk = sf_buf_kva(sf);
- bcopy(data, (void *)(sk + (base & PAGE_MASK)), size);
- shared_page_write_end(sf);
+ bcopy(data, shared_page_mapping + base, size);
}
static int
@@ -1596,6 +1569,7 @@
shared_page_init(void *dummy __unused)
{
vm_page_t m;
+ vm_offset_t addr;
sx_init(&shared_page_alloc_sx, "shpsx");
shared_page_obj = vm_pager_allocate(OBJT_PHYS, 0, PAGE_SIZE,
@@ -1605,15 +1579,95 @@
VM_ALLOC_ZERO);
m->valid = VM_PAGE_BITS_ALL;
VM_OBJECT_UNLOCK(shared_page_obj);
+ addr = kmem_alloc_nofault(kernel_map, PAGE_SIZE);
+ pmap_qenter(addr, &m, 1);
+ shared_page_mapping = (char *)addr;
}
SYSINIT(shp, SI_SUB_EXEC, SI_ORDER_FIRST, (sysinit_cfunc_t)shared_page_init,
NULL);
+static void
+timehands_update(struct sysentvec *sv)
+{
+ struct vdso_timehands th;
+ struct vdso_timekeep *tk;
+ uint32_t enabled, idx;
+
+ enabled = tc_fill_vdso_timehands(&th);
+ tk = (struct vdso_timekeep *)(shared_page_mapping +
+ sv->sv_timekeep_off);
+ idx = sv->sv_timekeep_curr;
+ atomic_store_rel_32(&tk->tk_th[idx].th_gen, 0);
+ if (++idx >= VDSO_TH_NUM)
+ idx = 0;
+ sv->sv_timekeep_curr = idx;
+ if (++sv->sv_timekeep_gen == 0)
+ sv->sv_timekeep_gen = 1;
+ th.th_gen = 0;
+ if (enabled)
+ tk->tk_th[idx] = th;
+ tk->tk_enabled = enabled;
+ atomic_store_rel_32(&tk->tk_th[idx].th_gen, sv->sv_timekeep_gen);
+ tk->tk_current = idx;
+}
+
+#ifdef COMPAT_FREEBSD32
+static void
+timehands_update32(struct sysentvec *sv)
+{
+ struct vdso_timekeep32 *tk;
+ struct vdso_timehands32 th;
+ uint32_t enabled, idx;
+
+ enabled = tc_fill_vdso_timehands32(&th);
+ tk = (struct vdso_timekeep32 *)(shared_page_mapping +
+ sv->sv_timekeep_off);
+ idx = sv->sv_timekeep_curr;
+ atomic_store_rel_32(&tk->tk_th[idx].th_gen, 0);
+ if (++idx >= VDSO_TH_NUM)
+ idx = 0;
+ sv->sv_timekeep_curr = idx;
+ if (++sv->sv_timekeep_gen == 0)
+ sv->sv_timekeep_gen = 1;
+ th.th_gen = 0;
+ if (enabled)
+ tk->tk_th[idx] = th;
+ tk->tk_enabled = enabled;
+ atomic_store_rel_32(&tk->tk_th[idx].th_gen, sv->sv_timekeep_gen);
+ tk->tk_current = idx;
+}
+#endif
+
+/*
+ * This is hackish, but easiest way to avoid creating list structures
+ * that needs to be iterated over from the hardclock interrupt
+ * context.
+ */
+static struct sysentvec *host_sysentvec;
+#ifdef COMPAT_FREEBSD32
+static struct sysentvec *compat32_sysentvec;
+#endif
+
void
+timekeep_push_vdso(void)
+{
+
+ if (host_sysentvec != NULL && host_sysentvec->sv_timekeep_base != 0)
+ timehands_update(host_sysentvec);
+#ifdef COMPAT_FREEBSD32
+ if (compat32_sysentvec != NULL &&
+ compat32_sysentvec->sv_timekeep_base != 0)
+ timehands_update32(compat32_sysentvec);
+#endif
+}
+
+void
exec_sysvec_init(void *param)
{
struct sysentvec *sv;
+ int tk_base;
+ uint32_t tk_ver;
sv = (struct sysentvec *)param;
@@ -1622,4 +1676,32 @@
sv->sv_shared_page_obj = shared_page_obj;
sv->sv_sigcode_base = sv->sv_shared_page_base +
shared_page_fill(*(sv->sv_szsigcode), 16, sv->sv_sigcode);
+ if ((sv->sv_flags & SV_ABI_MASK) != SV_ABI_FREEBSD)
+ return;
+ tk_ver = VDSO_TK_VER_CURR;
+#ifdef COMPAT_FREEBSD32
+ if ((sv->sv_flags & SV_ILP32) != 0) {
+ tk_base = shared_page_alloc(sizeof(struct vdso_timekeep32) +
+ sizeof(struct vdso_timehands32) * VDSO_TH_NUM, 16);
+ KASSERT(tk_base != -1, ("tk_base -1 for 32bit"));
+ shared_page_write(tk_base + offsetof(struct vdso_timekeep32,
+ tk_ver), sizeof(uint32_t), &tk_ver);
+ KASSERT(compat32_sysentvec == 0,
+ ("Native compat32 already registered"));
+ compat32_sysentvec = sv;
+ } else {
+#endif
+ tk_base = shared_page_alloc(sizeof(struct vdso_timekeep) +
+ sizeof(struct vdso_timehands) * VDSO_TH_NUM, 16);
+ KASSERT(tk_base != -1, ("tk_base -1 for native"));
+ shared_page_write(tk_base + offsetof(struct vdso_timekeep,
+ tk_ver), sizeof(uint32_t), &tk_ver);
+ KASSERT(host_sysentvec == 0, ("Native already registered"));
+ host_sysentvec = sv;
+#ifdef COMPAT_FREEBSD32
+ }
+#endif
+ sv->sv_timekeep_base = sv->sv_shared_page_base + tk_base;
+ sv->sv_timekeep_off = tk_base;
+ timekeep_push_vdso();
}
Modified: trunk/sys/sys/sysent.h
===================================================================
--- trunk/sys/sys/sysent.h 2016-09-15 22:56:13 UTC (rev 8098)
+++ trunk/sys/sys/sysent.h 2016-09-15 23:01:55 UTC (rev 8099)
@@ -124,6 +124,10 @@
vm_offset_t sv_shared_page_base;
vm_offset_t sv_shared_page_len;
vm_offset_t sv_sigcode_base;
+ vm_offset_t sv_timekeep_base;
+ int sv_timekeep_off;
+ int sv_timekeep_curr;
+ uint32_t sv_timekeep_gen;
void *sv_shared_page_obj;
void (*sv_schedtail)(struct thread *);
};
@@ -261,8 +265,6 @@
int shared_page_fill(int size, int align, const void *data);
void shared_page_write(int base, int size, const void *data);
void exec_sysvec_init(void *param);
-struct sf_buf *shared_page_write_start(int base);
-void shared_page_write_end(struct sf_buf *sf);
#define INIT_SYSENTVEC(name, sv) \
SYSINIT(name, SI_SUB_EXEC, SI_ORDER_ANY, \
Modified: trunk/sys/x86/x86/tsc.c
===================================================================
--- trunk/sys/x86/x86/tsc.c 2016-09-15 22:56:13 UTC (rev 8098)
+++ trunk/sys/x86/x86/tsc.c 2016-09-15 23:01:55 UTC (rev 8099)
@@ -27,6 +27,7 @@
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
+#include "opt_compat.h"
#include "opt_clock.h"
#include <sys/param.h>
@@ -41,6 +42,7 @@
#include <sys/kernel.h>
#include <sys/power.h>
#include <sys/smp.h>
+#include <sys/vdso.h>
#include <machine/clock.h>
#include <machine/cputypes.h>
#include <machine/md_var.h>
@@ -606,3 +608,23 @@
: "=a" (rv) : "c" ((int)(intptr_t)tc->tc_priv) : "edx");
return (rv);
}
+
+uint32_t
+cpu_fill_vdso_timehands(struct vdso_timehands *vdso_th)
+{
+
+ vdso_th->th_x86_shift = (int)(intptr_t)timecounter->tc_priv;
+ bzero(vdso_th->th_res, sizeof(vdso_th->th_res));
+ return (timecounter == &tsc_timecounter);
+}
+
+#ifdef COMPAT_FREEBSD32
+uint32_t
+cpu_fill_vdso_timehands32(struct vdso_timehands32 *vdso_th32)
+{
+
+ vdso_th32->th_x86_shift = (int)(intptr_t)timecounter->tc_priv;
+ bzero(vdso_th32->th_res, sizeof(vdso_th32->th_res));
+ return (timecounter == &tsc_timecounter);
+}
+#endif
More information about the Midnightbsd-cvs
mailing list