svn commit: r215661 - stable/8/sys/compat/linux
Author: netchild
Date: Mon Nov 22 08:21:58 2010
New Revision: 215661
URL: http://svn.freebsd.org/changeset/base/215661
Log:
MFC r215338:
- print out the PID and program name of the program trying to use an
unsupported futex operation
- for those futex operations which are known to be not supported,
print out which futex operation it is
- shortcut the error return of the unsupported FUTEX_CLOCK_REALTIME in
some cases:
FUTEX_CLOCK_REALTIME can be used to tell linux to use
CLOCK_REALTIME instead of CLOCK_MONOTONIC. FUTEX_CLOCK_REALTIME
however must only be set, if either FUTEX_WAIT_BITSET or
FUTEX_WAIT_REQUEUE_PI are set too. If that's not the case
we can die with ENOSYS right at the beginning.
Submitted by: arundel
Reviewed by:rdivacky (earlier iteration of the patch)
MFC after: 1 week
Modified:
stable/8/sys/compat/linux/linux_futex.c
stable/8/sys/compat/linux/linux_futex.h
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/compat/linux/linux_futex.c
==
--- stable/8/sys/compat/linux/linux_futex.c Mon Nov 22 08:19:18 2010
(r215660)
+++ stable/8/sys/compat/linux/linux_futex.c Mon Nov 22 08:21:58 2010
(r215661)
@@ -428,7 +428,7 @@ futex_atomic_op(struct thread *td, int e
int
linux_sys_futex(struct thread *td, struct linux_sys_futex_args *args)
{
- int op_ret, val, ret, nrwake;
+ int clockrt, nrwake, op_ret, ret, val;
struct linux_emuldata *em;
struct waiting_proc *wp;
struct futex *f, *f2;
@@ -441,7 +441,19 @@ linux_sys_futex(struct thread *td, struc
* in most cases (ie. when futexes are not shared on file descriptor
* or between different processes.).
*/
- args-op = (args-op ~LINUX_FUTEX_PRIVATE_FLAG);
+ args-op = args-op ~LINUX_FUTEX_PRIVATE_FLAG;
+
+ /*
+* Currently support for switching between CLOCK_MONOTONIC and
+* CLOCK_REALTIME is not present. However Linux forbids the use of
+* FUTEX_CLOCK_REALTIME with any op except FUTEX_WAIT_BITSET and
+* FUTEX_WAIT_REQUEUE_PI.
+*/
+ clockrt = args-op LINUX_FUTEX_CLOCK_REALTIME;
+ args-op = args-op ~LINUX_FUTEX_CLOCK_REALTIME;
+ if (clockrt args-op != LINUX_FUTEX_WAIT_BITSET
+ args-op != LINUX_FUTEX_WAIT_REQUEUE_PI)
+ return (ENOSYS);
switch (args-op) {
case LINUX_FUTEX_WAIT:
@@ -624,14 +636,23 @@ linux_sys_futex(struct thread *td, struc
case LINUX_FUTEX_LOCK_PI:
/* not yet implemented */
+ linux_msg(td,
+ linux_sys_futex:
+ op LINUX_FUTEX_LOCK_PI not implemented.\n);
return (ENOSYS);
case LINUX_FUTEX_UNLOCK_PI:
/* not yet implemented */
+ linux_msg(td,
+ linux_sys_futex:
+ op LINUX_FUTEX_UNLOCK_PI not implemented.\n);
return (ENOSYS);
case LINUX_FUTEX_TRYLOCK_PI:
/* not yet implemented */
+ linux_msg(td,
+ linux_sys_futex:
+ op LINUX_FUTEX_TRYLOCK_PI not implemented.\n);
return (ENOSYS);
case LINUX_FUTEX_REQUEUE:
@@ -644,15 +665,30 @@ linux_sys_futex(struct thread *td, struc
*/
em = em_find(td-td_proc, EMUL_DONTLOCK);
if (em-used_requeue == 0) {
- printf(linux(%s (%d)) sys_futex:
- unsupported futex_requeue op\n,
- td-td_proc-p_comm, td-td_proc-p_pid);
+ linux_msg(td,
+ linux_sys_futex:
+ unsupported futex_requeue op\n);
em-used_requeue = 1;
}
return (EINVAL);
+ case LINUX_FUTEX_WAIT_BITSET:
+ /* not yet implemented */
+ linux_msg(td,
+ linux_sys_futex:
+ op FUTEX_WAIT_BITSET not implemented.\n);
+ return (ENOSYS);
+
+ case LINUX_FUTEX_WAIT_REQUEUE_PI:
+ /* not yet implemented */
+ linux_msg(td,
+ linux_sys_futex:
+ op FUTEX_WAIT_REQUEUE_PI not implemented.\n);
+ return (ENOSYS);
+
default:
- printf(linux_sys_futex: unknown op %d\n, args-op);
+ linux_msg(td,
+
svn commit: r215662 - stable/8/sys/compat/linux
Author: netchild
Date: Mon Nov 22 08:35:06 2010
New Revision: 215662
URL: http://svn.freebsd.org/changeset/base/215662
Log:
MFC r215339:
Some style(9) fixes.
Submitted by: arundel
Modified:
stable/8/sys/compat/linux/linux_futex.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/compat/linux/linux_futex.c
==
--- stable/8/sys/compat/linux/linux_futex.c Mon Nov 22 08:21:58 2010
(r215661)
+++ stable/8/sys/compat/linux/linux_futex.c Mon Nov 22 08:35:06 2010
(r215662)
@@ -668,7 +668,7 @@ linux_sys_futex(struct thread *td, struc
linux_msg(td,
linux_sys_futex:
unsupported futex_requeue op\n);
- em-used_requeue = 1;
+ em-used_requeue = 1;
}
return (EINVAL);
@@ -713,7 +713,7 @@ linux_set_robust_list(struct thread *td,
em-robust_futexes = args-head;
EMUL_UNLOCK(emul_lock);
- return (0);
+ return (0);
}
int
@@ -731,7 +731,7 @@ linux_get_robust_list(struct thread *td,
if (!args-pid) {
em = em_find(td-td_proc, EMUL_DONTLOCK);
- head = em-robust_futexes;
+ head = em-robust_futexes;
} else {
struct proc *p;
@@ -741,14 +741,14 @@ linux_get_robust_list(struct thread *td,
em = em_find(p, EMUL_DONTLOCK);
/* XXX: ptrace? */
- if (priv_check(td, PRIV_CRED_SETUID) ||
+ if (priv_check(td, PRIV_CRED_SETUID) ||
priv_check(td, PRIV_CRED_SETEUID) ||
p_candebug(td, p)) {
PROC_UNLOCK(p);
return (EPERM);
}
head = em-robust_futexes;
-
+
PROC_UNLOCK(p);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215663 - head/sys/i386/xen
Author: cperciva Date: Mon Nov 22 09:04:29 2010 New Revision: 215663 URL: http://svn.freebsd.org/changeset/base/215663 Log: In xen_get_timecount, return the full ns-precision time rather than rounding to 1/HZ precision. I have no idea why the rounding was introduced in the first place, but it makes FreeBSD unhappy. Modified: head/sys/i386/xen/clock.c Modified: head/sys/i386/xen/clock.c == --- head/sys/i386/xen/clock.c Mon Nov 22 08:35:06 2010(r215662) +++ head/sys/i386/xen/clock.c Mon Nov 22 09:04:29 2010(r215663) @@ -829,7 +829,7 @@ xen_get_timecount(struct timecounter *tc clk = shadow-system_timestamp + get_nsec_offset(shadow); - return (uint32_t)((clk / NS_PER_TICK) * NS_PER_TICK); + return (uint32_t)(clk); } ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215664 - in head/sys: compat/linux kern
Author: netchild
Date: Mon Nov 22 09:06:59 2010
New Revision: 215664
URL: http://svn.freebsd.org/changeset/base/215664
Log:
By using the 32-bit Linux version of Sun's Java Development Kit 1.6
on FreeBSD (amd64), invocations of javac (or java) eventually
end with the output of Killed and exit code 137.
This is caused by:
1. After calling exec() in multithreaded linux program threads are not
destroyed and continue running. They get killed after program being
executed finishes.
2. linux_exit_group doesn't return correct exit code when called not
from group leader. Which happens regularly using sun jvm.
The submitters fix this in a similar way to how NetBSD handles this.
I took the PRs away from dchagin, who seems to be out of touch of
this since a while (no response from him).
The patches committed here are from [2], with some little modifications
from me to the style.
PR: 141439 [1], 144194 [2]
Submitted by: Stefan Schmidt stefan.schm...@stadtbuch.de, gk
Reviewed by: rdivacky (in april 2010)
MFC after:5 days
Modified:
head/sys/compat/linux/linux_emul.c
head/sys/compat/linux/linux_emul.h
head/sys/compat/linux/linux_misc.c
head/sys/kern/kern_exit.c
Modified: head/sys/compat/linux/linux_emul.c
==
--- head/sys/compat/linux/linux_emul.c Mon Nov 22 09:04:29 2010
(r215663)
+++ head/sys/compat/linux/linux_emul.c Mon Nov 22 09:06:59 2010
(r215664)
@@ -155,7 +155,7 @@ void
linux_proc_exit(void *arg __unused, struct proc *p)
{
struct linux_emuldata *em;
- int error;
+ int error, shared_flags, shared_xstat;
struct thread *td = FIRST_THREAD_IN_PROC(p);
int *child_clear_tid;
struct proc *q, *nq;
@@ -187,6 +187,8 @@ linux_proc_exit(void *arg __unused, stru
}
EMUL_SHARED_WLOCK(emul_shared_lock);
+ shared_flags = em-shared-flags;
+ shared_xstat = em-shared-xstat;
LIST_REMOVE(em, threads);
em-shared-refs--;
@@ -196,6 +198,12 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
+ PROC_LOCK(p);
+ p-p_xstat = shared_xstat;
+ PROC_UNLOCK(p);
+ }
+
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
int null = 0;
@@ -257,6 +265,9 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
+ if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ /* Kill threads regardless of imgp-sysent value */
+ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
if (__predict_false(imgp-sysent != elf_linux_sysvec
p-p_sysent == elf_linux_sysvec)) {
struct linux_emuldata *em;
@@ -334,3 +345,29 @@ linux_set_tid_address(struct thread *td,
EMUL_UNLOCK(emul_lock);
return 0;
}
+
+void
+linux_kill_threads(struct thread *td, int sig)
+{
+ struct linux_emuldata *em, *td_em, *tmp_em;
+ struct proc *sp;
+
+ td_em = em_find(td-td_proc, EMUL_DONTLOCK);
+
+ KASSERT(td_em != NULL, (linux_kill_threads: emuldata not found.\n));
+
+ EMUL_SHARED_RLOCK(emul_shared_lock);
+ LIST_FOREACH_SAFE(em, td_em-shared-threads, threads, tmp_em) {
+ if (em-pid == td_em-pid)
+ continue;
+
+ sp = pfind(em-pid);
+ if ((sp-p_flag P_WEXIT) == 0)
+ psignal(sp, sig);
+ PROC_UNLOCK(sp);
+#ifdef DEBUG
+ printf(LMSG(linux_kill_threads: kill PID %d\n), em-pid);
+#endif
+ }
+ EMUL_SHARED_RUNLOCK(emul_shared_lock);
+}
Modified: head/sys/compat/linux/linux_emul.h
==
--- head/sys/compat/linux/linux_emul.h Mon Nov 22 09:04:29 2010
(r215663)
+++ head/sys/compat/linux/linux_emul.h Mon Nov 22 09:06:59 2010
(r215664)
@@ -31,8 +31,12 @@
#ifndef _LINUX_EMUL_H_
#define_LINUX_EMUL_H_
+#define EMUL_SHARED_HASXSTAT 0x01
+
struct linux_emuldata_shared {
int refs;
+ int flags;
+ int xstat;
pid_t group_pid;
LIST_HEAD(, linux_emuldata) threads; /* head of list of linux threads */
@@ -76,6 +80,7 @@ int linux_proc_init(struct thread *, pid
void linux_proc_exit(void *, struct proc *);
void linux_schedtail(void *, struct proc *);
void linux_proc_exec(void *, struct proc *, struct image_params *);
+void linux_kill_threads(struct thread *, int);
extern struct sx emul_shared_lock;
extern struct mtx emul_lock;
Modified:
svn commit: r215665 - head/sys/kern
Author: cperciva
Date: Mon Nov 22 09:13:25 2010
New Revision: 215665
URL: http://svn.freebsd.org/changeset/base/215665
Log:
In tc_windup, handle the case where the previous call to tc_windup was
more than 1s earlier. Prior to this commit, the computation of
th_scale * delta (which produces a 64-bit value equal to the time since
the last tc_windup call in units of 2^(-64) seconds) would overflow and
any complete seconds would be lost.
We fix this by repeatedly converting tc_frequency units of timecounter
to one seconds; this is not exactly correct, since it loses the NTP
adjustment, but if we find ourselves going more than 1s at a time between
clock interrupts, losing a few seconds worth of NTP adjustments is the
least of our problems...
Modified:
head/sys/kern/kern_tc.c
Modified: head/sys/kern/kern_tc.c
==
--- head/sys/kern/kern_tc.c Mon Nov 22 09:06:59 2010(r215664)
+++ head/sys/kern/kern_tc.c Mon Nov 22 09:13:25 2010(r215665)
@@ -442,6 +442,16 @@ tc_windup(void)
ncount = 0;
th-th_offset_count += delta;
th-th_offset_count = th-th_counter-tc_counter_mask;
+ while (delta th-th_counter-tc_frequency) {
+ /* Eat complete unadjusted seconds. */
+ delta -= th-th_counter-tc_frequency;
+ th-th_offset.sec++;
+ }
+ if ((delta th-th_counter-tc_frequency / 2)
+ (th-th_scale * delta (uint64_t)1 63)) {
+ /* The product th_scale * delta just barely overflows. */
+ th-th_offset.sec++;
+ }
bintime_addx(th-th_offset, th-th_scale * delta);
/*
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215666 - head/sys/compat/linux
Author: netchild Date: Mon Nov 22 09:25:32 2010 New Revision: 215666 URL: http://svn.freebsd.org/changeset/base/215666 Log: Remove trailing dot from the unimplemented futex messages to make them consistent with the syscall and ipc messages. Submitted by: arundel MFC after:3 days Modified: head/sys/compat/linux/linux_futex.c Modified: head/sys/compat/linux/linux_futex.c == --- head/sys/compat/linux/linux_futex.c Mon Nov 22 09:13:25 2010 (r215665) +++ head/sys/compat/linux/linux_futex.c Mon Nov 22 09:25:32 2010 (r215666) @@ -626,21 +626,21 @@ linux_sys_futex(struct thread *td, struc /* not yet implemented */ linux_msg(td, linux_sys_futex: - op LINUX_FUTEX_LOCK_PI not implemented.\n); + op LINUX_FUTEX_LOCK_PI not implemented\n); return (ENOSYS); case LINUX_FUTEX_UNLOCK_PI: /* not yet implemented */ linux_msg(td, linux_sys_futex: - op LINUX_FUTEX_UNLOCK_PI not implemented.\n); + op LINUX_FUTEX_UNLOCK_PI not implemented\n); return (ENOSYS); case LINUX_FUTEX_TRYLOCK_PI: /* not yet implemented */ linux_msg(td, linux_sys_futex: - op LINUX_FUTEX_TRYLOCK_PI not implemented.\n); + op LINUX_FUTEX_TRYLOCK_PI not implemented\n); return (ENOSYS); case LINUX_FUTEX_REQUEUE: @@ -664,14 +664,14 @@ linux_sys_futex(struct thread *td, struc /* not yet implemented */ linux_msg(td, linux_sys_futex: - op FUTEX_WAIT_BITSET not implemented.\n); + op FUTEX_WAIT_BITSET not implemented\n); return (ENOSYS); case LINUX_FUTEX_WAIT_REQUEUE_PI: /* not yet implemented */ linux_msg(td, linux_sys_futex: - op FUTEX_WAIT_REQUEUE_PI not implemented.\n); + op FUTEX_WAIT_REQUEUE_PI not implemented\n); return (ENOSYS); default: ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215664 - in head/sys: compat/linux kern
On Mon, Nov 22, 2010 at 09:07:00AM +, Alexander Leidinger wrote:
Author: netchild
Date: Mon Nov 22 09:06:59 2010
New Revision: 215664
URL: http://svn.freebsd.org/changeset/base/215664
Log:
By using the 32-bit Linux version of Sun's Java Development Kit 1.6
on FreeBSD (amd64), invocations of javac (or java) eventually
end with the output of Killed and exit code 137.
This is caused by:
1. After calling exec() in multithreaded linux program threads are not
destroyed and continue running. They get killed after program being
executed finishes.
2. linux_exit_group doesn't return correct exit code when called not
from group leader. Which happens regularly using sun jvm.
The submitters fix this in a similar way to how NetBSD handles this.
I took the PRs away from dchagin, who seems to be out of touch of
this since a while (no response from him).
The patches committed here are from [2], with some little modifications
from me to the style.
PR: 141439 [1], 144194 [2]
Submitted by: Stefan Schmidt stefan.schm...@stadtbuch.de, gk
Reviewed by:rdivacky (in april 2010)
MFC after: 5 days
Modified:
head/sys/compat/linux/linux_emul.c
head/sys/compat/linux/linux_emul.h
head/sys/compat/linux/linux_misc.c
head/sys/kern/kern_exit.c
Modified: head/sys/compat/linux/linux_emul.c
==
--- head/sys/compat/linux/linux_emul.cMon Nov 22 09:04:29 2010
(r215663)
+++ head/sys/compat/linux/linux_emul.cMon Nov 22 09:06:59 2010
(r215664)
@@ -155,7 +155,7 @@ void
linux_proc_exit(void *arg __unused, struct proc *p)
{
struct linux_emuldata *em;
- int error;
+ int error, shared_flags, shared_xstat;
struct thread *td = FIRST_THREAD_IN_PROC(p);
int *child_clear_tid;
struct proc *q, *nq;
@@ -187,6 +187,8 @@ linux_proc_exit(void *arg __unused, stru
}
EMUL_SHARED_WLOCK(emul_shared_lock);
+ shared_flags = em-shared-flags;
+ shared_xstat = em-shared-xstat;
LIST_REMOVE(em, threads);
em-shared-refs--;
@@ -196,6 +198,12 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
+ PROC_LOCK(p);
+ p-p_xstat = shared_xstat;
+ PROC_UNLOCK(p);
+ }
Why is process lock taken there ? The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
+
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
int null = 0;
@@ -257,6 +265,9 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
+ if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ /* Kill threads regardless of imgp-sysent value */
+ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
This is better expressed by
if ((p-p_sysent-sv_flags SV_ABI_MASK) == SV_ABI_LINUX)
Regardless of this mostly cosmetic issue, this is racy. Other
linux thread in the same process might do an execve(3).
More, if execve(3) call fails, then you return into the process
that lacks all threads except the one that called execve(3).
if (__predict_false(imgp-sysent != elf_linux_sysvec
p-p_sysent == elf_linux_sysvec)) {
struct linux_emuldata *em;
@@ -334,3 +345,29 @@ linux_set_tid_address(struct thread *td,
EMUL_UNLOCK(emul_lock);
return 0;
}
+
+void
+linux_kill_threads(struct thread *td, int sig)
+{
+ struct linux_emuldata *em, *td_em, *tmp_em;
+ struct proc *sp;
+
+ td_em = em_find(td-td_proc, EMUL_DONTLOCK);
+
+ KASSERT(td_em != NULL, (linux_kill_threads: emuldata not found.\n));
+
+ EMUL_SHARED_RLOCK(emul_shared_lock);
+ LIST_FOREACH_SAFE(em, td_em-shared-threads, threads, tmp_em) {
+ if (em-pid == td_em-pid)
+ continue;
+
+ sp = pfind(em-pid);
+ if ((sp-p_flag P_WEXIT) == 0)
+ psignal(sp, sig);
+ PROC_UNLOCK(sp);
+#ifdef DEBUG
+ printf(LMSG(linux_kill_threads: kill PID %d\n), em-pid);
+#endif
+ }
+ EMUL_SHARED_RUNLOCK(emul_shared_lock);
+}
Modified: head/sys/compat/linux/linux_emul.h
==
--- head/sys/compat/linux/linux_emul.hMon Nov 22 09:04:29 2010
(r215663)
+++ head/sys/compat/linux/linux_emul.hMon Nov 22 09:06:59 2010
(r215664)
@@
svn commit: r215667 - stable/8/sbin/fdisk
Author: brian Date: Mon Nov 22 09:32:54 2010 New Revision: 215667 URL: http://svn.freebsd.org/changeset/base/215667 Log: MFC r212247 r212724 from head: Handle geli-encrypted root disk devices. Add support for identifying a journaled root filesystem. Fix support for identifying the given /dev/vinum/root example. Modified: stable/8/sbin/fdisk/fdisk.c Directory Properties: stable/8/sbin/fdisk/ (props changed) Modified: stable/8/sbin/fdisk/fdisk.c == --- stable/8/sbin/fdisk/fdisk.c Mon Nov 22 09:25:32 2010(r215666) +++ stable/8/sbin/fdisk/fdisk.c Mon Nov 22 09:32:54 2010(r215667) @@ -1461,6 +1461,8 @@ sanitize_partition(struct dos_partition * /dev/ad0s1a = /dev/ad0 * /dev/da0a = /dev/da0 * /dev/vinum/root = /dev/vinum/root + * A .eli part is removed if it exists (see geli(8)). + * A .journal ending is removed if it exists (see gjournal(8)). */ static char * get_rootdisk(void) @@ -1469,16 +1471,20 @@ get_rootdisk(void) regex_t re; #define NMATCHES 2 regmatch_t rm[NMATCHES]; - char *s; + char dev[PATH_MAX], *s; int rv; if (statfs(/, rootfs) == -1) err(1, statfs(\/\)); - if ((rv = regcomp(re, ^(/dev/[a-z/]+[0-9]+)([sp][0-9]+)?[a-h]?$, + if ((rv = regcomp(re, ^(/dev/[a-z/]+[0-9]*)([sp][0-9]+)?[a-h]?(\\.journal)?$, REG_EXTENDED)) != 0) errx(1, regcomp() failed (%d), rv); - if ((rv = regexec(re, rootfs.f_mntfromname, NMATCHES, rm, 0)) != 0) + strlcpy(dev, rootfs.f_mntfromname, sizeof (dev)); + if ((s = strstr(dev, .eli)) != NULL) + memmove(s, s+4, strlen(s + 4) + 1); + + if ((rv = regexec(re, dev, NMATCHES, rm, 0)) != 0) errx(1, mounted root fs resource doesn't match expectations (regexec returned %d), rv); ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215668 - stable/7/sbin/fdisk
Author: brian Date: Mon Nov 22 09:34:38 2010 New Revision: 215668 URL: http://svn.freebsd.org/changeset/base/215668 Log: MFC r212247 r212724 from head: Handle geli-encrypted root disk devices. Add support for identifying a journaled root filesystem. Fix support for identifying the given /dev/vinum/root example. Modified: stable/7/sbin/fdisk/fdisk.c Directory Properties: stable/7/sbin/fdisk/ (props changed) Modified: stable/7/sbin/fdisk/fdisk.c == --- stable/7/sbin/fdisk/fdisk.c Mon Nov 22 09:32:54 2010(r215667) +++ stable/7/sbin/fdisk/fdisk.c Mon Nov 22 09:34:38 2010(r215668) @@ -1492,6 +1492,8 @@ sanitize_partition(struct dos_partition * /dev/ad0s1a = /dev/ad0 * /dev/da0a = /dev/da0 * /dev/vinum/root = /dev/vinum/root + * A .eli part is removed if it exists (see geli(8)). + * A .journal ending is removed if it exists (see gjournal(8)). */ static char * get_rootdisk(void) @@ -1500,16 +1502,20 @@ get_rootdisk(void) regex_t re; #define NMATCHES 2 regmatch_t rm[NMATCHES]; - char *s; + char dev[PATH_MAX], *s; int rv; if (statfs(/, rootfs) == -1) err(1, statfs(\/\)); - if ((rv = regcomp(re, ^(/dev/[a-z/]+[0-9]+)([sp][0-9]+)?[a-h]?$, + if ((rv = regcomp(re, ^(/dev/[a-z/]+[0-9]*)([sp][0-9]+)?[a-h]?(\\.journal)?$, REG_EXTENDED)) != 0) errx(1, regcomp() failed (%d), rv); - if ((rv = regexec(re, rootfs.f_mntfromname, NMATCHES, rm, 0)) != 0) + strlcpy(dev, rootfs.f_mntfromname, sizeof (dev)); + if ((s = strstr(dev, .eli)) != NULL) + memmove(s, s+4, strlen(s + 4) + 1); + + if ((rv = regexec(re, dev, NMATCHES, rm, 0)) != 0) errx(1, mounted root fs resource doesn't match expectations (regexec returned %d), rv); ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215669 - head
Author: netchild Date: Mon Nov 22 09:37:52 2010 New Revision: 215669 URL: http://svn.freebsd.org/changeset/base/215669 Log: 1) Add a hint to check for duplicates with optional files. The committed text is a little bit modified to what was submitted. The code example to automate a part of this was proposed by Dmitry Morozovsky. 2) Remove trailing whitespace. Submitted by: arundel Modified: head/ObsoleteFiles.inc Modified: head/ObsoleteFiles.inc == --- head/ObsoleteFiles.inc Mon Nov 22 09:34:38 2010(r215668) +++ head/ObsoleteFiles.inc Mon Nov 22 09:37:52 2010(r215669) @@ -13,6 +13,15 @@ # # The file is partitioned: OLD_FILES first, then OLD_LIBS and OLD_DIRS last. # +# Before you commit changes to this file please check if any entries in +# tools/build/mk/OptionalObsoleteFiles.inc can be removed. The following +# command tells which files are listed more than once regardless of some +# architecture specific conditionals, so you can not blindly trust the +# output: +# ( grep '+=' /usr/src/ObsoleteFiles.inc | sort -u ; \ +# grep '+=' /usr/src/tools/build/mk/OptionalObsoleteFiles.inc | sort -u) | \ +# sort | uniq -d +# # 20101112: vgonel(9) has gone to private API a while ago OLD_FILES+=usr/share/man/man9/vgonel.9.gz @@ -1634,7 +1643,7 @@ OLD_DIRS+=usr/include/c++/3.4 OLD_FILES+=usr/sbin/zfs OLD_FILES+=usr/sbin/zpool # 20070423: rc.bluetooth (examples) removed -OLD_FILES+=usr/share/examples/netgraph/bluetooth/rc.bluetooth +OLD_FILES+=usr/share/examples/netgraph/bluetooth/rc.bluetooth # 20070421: worm.4 removed OLD_FILES+=usr/share/man/man4/worm.4.gz # 20070417: trunk(4) renamed to lagg(4) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215670 - svnadmin/conf
Author: mav Date: Mon Nov 22 09:59:10 2010 New Revision: 215670 URL: http://svn.freebsd.org/changeset/base/215670 Log: Release ae@ from mentorship. He is now ready wear his own pointy hats. Congratulations! Reviewed by: kib@ Modified: svnadmin/conf/mentors Modified: svnadmin/conf/mentors == --- svnadmin/conf/mentors Mon Nov 22 09:37:52 2010(r215669) +++ svnadmin/conf/mentors Mon Nov 22 09:59:10 2010(r215670) @@ -10,7 +10,6 @@ # Sort by mentee login name. # Mentee Mentor Optional comment -ae mav Co-mentor: kib anchie bz andreast nwhitehorn andrew imp ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215671 - head/sbin/geom/class/part
Author: ae
Date: Mon Nov 22 10:08:33 2010
New Revision: 215671
URL: http://svn.freebsd.org/changeset/base/215671
Log:
Always dump partition labels with `gpart backup`, but `gpart restore`
does restore them only when -l option is specified [1]. Make number of
entries field in backup format optional. Document -l and -r options of
`gpart show` action.
Suggested by: pjd [1]
MFC after:1 week
Modified:
head/sbin/geom/class/part/geom_part.c
head/sbin/geom/class/part/gpart.8
Modified: head/sbin/geom/class/part/geom_part.c
==
--- head/sbin/geom/class/part/geom_part.c Mon Nov 22 09:59:10 2010
(r215670)
+++ head/sbin/geom/class/part/geom_part.c Mon Nov 22 10:08:33 2010
(r215671)
@@ -100,10 +100,8 @@ struct g_command PUBSYM(class_commands)[
G_OPT_SENTINEL },
[-b start] [-s size] -t type [-i index] [-l label] [-f flags] geom
},
- { backup, 0, gpart_backup, {
- { 'l', backup_labels, NULL, G_TYPE_BOOL},
- G_OPT_SENTINEL },
- [-l] geom
+ { backup, 0, gpart_backup, G_NULL_OPTS,
+ geom
},
{ bootcode, 0, gpart_bootcode, {
{ 'b', GPART_PARAM_BOOTCODE, G_VAL_OPTIONAL, G_TYPE_STRING },
@@ -175,9 +173,10 @@ struct g_command PUBSYM(class_commands)[
},
{ restore, 0, gpart_restore, {
{ 'F', force, NULL, G_TYPE_BOOL },
+ { 'l', restore_labels, NULL, G_TYPE_BOOL },
{ 'f', flags, GPART_FLAGS, G_TYPE_STRING },
G_OPT_SENTINEL },
- [-F] [-f flags] provider [...]
+ [-lF] [-f flags] provider [...]
},
{ recover, 0, gpart_issue, {
{ 'f', flags, GPART_FLAGS, G_TYPE_STRING },
@@ -678,7 +677,7 @@ gpart_backup(struct gctl_req *req, unsig
const char *s, *scheme;
off_t sector, end;
off_t length, secsz;
- int error, labels, i, windex, wblocks, wtype;
+ int error, i, windex, wblocks, wtype;
if (gctl_get_int(req, nargs) != 1)
errx(EXIT_FAILURE, Invalid number of arguments.);
@@ -696,7 +695,6 @@ gpart_backup(struct gctl_req *req, unsig
s = gctl_get_ascii(req, arg0);
if (s == NULL)
abort();
- labels = gctl_get_int(req, backup_labels);
gp = find_geom(classp, s);
if (gp == NULL)
errx(EXIT_FAILURE, No such geom: %s., s);
@@ -734,14 +732,12 @@ gpart_backup(struct gctl_req *req, unsig
length = end - sector + 1;
}
s = find_provcfg(pp, label);
- printf(%-*s %*s %*jd %*jd,
+ printf(%-*s %*s %*jd %*jd %s %s\n,
windex, find_provcfg(pp, index),
wtype, find_provcfg(pp, type),
wblocks, (intmax_t)sector,
- wblocks, (intmax_t)length);
- if (labels s != NULL)
- printf( %s, s);
- printf( %s\n, fmtattrib(pp));
+ wblocks, (intmax_t)length,
+ (s != NULL) ? s: , fmtattrib(pp));
}
geom_deletetree(mesh);
}
@@ -769,7 +765,7 @@ gpart_restore(struct gctl_req *req, unsi
struct ggeom *gp;
const char *s, *flags, *errstr, *label;
char **ap, *argv[6], line[BUFSIZ], *pline;
- int error, forced, i, l, nargs, created;
+ int error, forced, i, l, nargs, created, rl;
intmax_t n;
nargs = gctl_get_int(req, nargs);
@@ -778,6 +774,7 @@ gpart_restore(struct gctl_req *req, unsi
forced = gctl_get_int(req, force);
flags = gctl_get_ascii(req, flags);
+ rl = gctl_get_int(req, restore_labels);
s = gctl_get_ascii(req, class);
if (s == NULL)
abort();
@@ -829,19 +826,21 @@ gpart_restore(struct gctl_req *req, unsi
break;
l = ap - argv[0];
label = pline = NULL;
- if (l == 2) { /* create table */
+ if (l == 1 || l == 2) { /* create table */
if (created)
errx(EXIT_FAILURE, Incorrect backup format.);
- n = atoi(argv[1]);
+ if (l == 2)
+ n = strtoimax(argv[1], NULL, 0);
for (i = 0; i nargs; i++) {
s = gctl_get_ascii(req, arg%d, i);
r = gctl_get_handle();
- n = strtoimax(argv[1], NULL, 0);
gctl_ro_param(r, class, -1,
classp-lg_name);
gctl_ro_param(r, verb, -1, create);
gctl_ro_param(r, scheme, -1, argv[0]);
- gctl_ro_param(r, entries,
Re: svn commit: r215664 - in head/sys: compat/linux kern
Quoting Kostik Belousov kostik...@gmail.com (from Mon, 22 Nov 2010
11:31:34 +0200):
On Mon, Nov 22, 2010 at 09:07:00AM +, Alexander Leidinger wrote:
Author: netchild
Date: Mon Nov 22 09:06:59 2010
New Revision: 215664
URL: http://svn.freebsd.org/changeset/base/215664
Log:
By using the 32-bit Linux version of Sun's Java Development Kit 1.6
on FreeBSD (amd64), invocations of javac (or java) eventually
end with the output of Killed and exit code 137.
@@ -196,6 +198,12 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
+ PROC_LOCK(p);
+ p-p_xstat = shared_xstat;
+ PROC_UNLOCK(p);
+ }
Why is process lock taken there ? The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
Change below.
+
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
int null = 0;
@@ -257,6 +265,9 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
+ if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ /* Kill threads regardless of imgp-sysent value */
+ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
This is better expressed by
if ((p-p_sysent-sv_flags SV_ABI_MASK) == SV_ABI_LINUX)
Is this OK for you?
---snip---
Index: compat/linux/linux_emul.c
===
--- compat/linux/linux_emul.c (Revision 215664)
+++ compat/linux/linux_emul.c (Arbeitskopie)
@@ -198,11 +198,8 @@
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
- if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
- PROC_LOCK(p);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0)
p-p_xstat = shared_xstat;
- PROC_UNLOCK(p);
- }
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
@@ -265,7 +262,8 @@
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
- if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) ==
+ SV_ABI_LINUX))
/* Kill threads regardless of imgp-sysent value */
linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
if (__predict_false(imgp-sysent != elf_linux_sysvec
---snip---
Regardless of this mostly cosmetic issue, this is racy. Other
linux thread in the same process might do an execve(3).
More, if execve(3) call fails, then you return into the process
that lacks all threads except the one that called execve(3).
How critical is this in your opinion (relative to the issue this patch
is fixing)? Do you prefer a backout or do you think the probability
that the someone wins the race is low enough?
Do you see a solution for the race?
Bye,
Alexander.
--
http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID = B0063FE7
http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID = 72077137
This generation doesn't have emotional baggage.
We have emotional moving vans.
-- Bruce Feirstein
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215664 - in head/sys: compat/linux kern
On Mon, Nov 22, 2010 at 11:13:06AM +0100, Alexander Leidinger wrote:
Quoting Kostik Belousov kostik...@gmail.com (from Mon, 22 Nov 2010
11:31:34 +0200):
On Mon, Nov 22, 2010 at 09:07:00AM +, Alexander Leidinger wrote:
Author: netchild
Date: Mon Nov 22 09:06:59 2010
New Revision: 215664
URL: http://svn.freebsd.org/changeset/base/215664
Log:
By using the 32-bit Linux version of Sun's Java Development Kit 1.6
on FreeBSD (amd64), invocations of javac (or java) eventually
end with the output of Killed and exit code 137.
@@ -196,6 +198,12 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
+ PROC_LOCK(p);
+ p-p_xstat = shared_xstat;
+ PROC_UNLOCK(p);
+ }
Why is process lock taken there ? The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
Change below.
+
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
int null = 0;
@@ -257,6 +265,9 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
+ if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ /* Kill threads regardless of imgp-sysent value */
+ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
This is better expressed by
if ((p-p_sysent-sv_flags SV_ABI_MASK) == SV_ABI_LINUX)
Is this OK for you?
---snip---
Index: compat/linux/linux_emul.c
===
--- compat/linux/linux_emul.c (Revision 215664)
+++ compat/linux/linux_emul.c (Arbeitskopie)
@@ -198,11 +198,8 @@
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
- if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
- PROC_LOCK(p);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0)
p-p_xstat = shared_xstat;
- PROC_UNLOCK(p);
- }
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
@@ -265,7 +262,8 @@
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
- if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) ==
+ SV_ABI_LINUX))
/* Kill threads regardless of imgp-sysent value */
linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
if (__predict_false(imgp-sysent != elf_linux_sysvec
---snip---
Yes.
Regardless of this mostly cosmetic issue, this is racy. Other
linux thread in the same process might do an execve(3).
More, if execve(3) call fails, then you return into the process
that lacks all threads except the one that called execve(3).
How critical is this in your opinion (relative to the issue this patch
is fixing)? Do you prefer a backout or do you think the probability
that the someone wins the race is low enough?
Do you see a solution for the race?
I did not asked for backout, nor I am asking now.
Most likely, the semantic of linux thread groups cannot be implemented
by only using event handlers that linux.ko hooks now. How linux handles
single-threading when doing execve(2) from multithreaded process ?
pgp33400VCDnC.pgp
Description: PGP signature
Re: svn commit: r215664 - in head/sys: compat/linux kern
On (22/11/2010 11:31), Kostik Belousov wrote:
On Mon, Nov 22, 2010 at 09:07:00AM +, Alexander Leidinger wrote:
Author: netchild
Date: Mon Nov 22 09:06:59 2010
New Revision: 215664
URL: http://svn.freebsd.org/changeset/base/215664
Log:
By using the 32-bit Linux version of Sun's Java Development Kit 1.6
on FreeBSD (amd64), invocations of javac (or java) eventually
end with the output of Killed and exit code 137.
This is caused by:
1. After calling exec() in multithreaded linux program threads are not
destroyed and continue running. They get killed after program being
executed finishes.
2. linux_exit_group doesn't return correct exit code when called not
from group leader. Which happens regularly using sun jvm.
The submitters fix this in a similar way to how NetBSD handles this.
I took the PRs away from dchagin, who seems to be out of touch of
this since a while (no response from him).
The patches committed here are from [2], with some little modifications
from me to the style.
PR: 141439 [1], 144194 [2]
Submitted by: Stefan Schmidt stefan.schm...@stadtbuch.de, gk
Reviewed by: rdivacky (in april 2010)
MFC after:5 days
Modified:
head/sys/compat/linux/linux_emul.c
head/sys/compat/linux/linux_emul.h
head/sys/compat/linux/linux_misc.c
head/sys/kern/kern_exit.c
Modified: head/sys/compat/linux/linux_emul.c
==
--- head/sys/compat/linux/linux_emul.c Mon Nov 22 09:04:29 2010
(r215663)
+++ head/sys/compat/linux/linux_emul.c Mon Nov 22 09:06:59 2010
(r215664)
@@ -155,7 +155,7 @@ void
linux_proc_exit(void *arg __unused, struct proc *p)
{
struct linux_emuldata *em;
- int error;
+ int error, shared_flags, shared_xstat;
struct thread *td = FIRST_THREAD_IN_PROC(p);
int *child_clear_tid;
struct proc *q, *nq;
@@ -187,6 +187,8 @@ linux_proc_exit(void *arg __unused, stru
}
EMUL_SHARED_WLOCK(emul_shared_lock);
+ shared_flags = em-shared-flags;
+ shared_xstat = em-shared-xstat;
LIST_REMOVE(em, threads);
em-shared-refs--;
@@ -196,6 +198,12 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
+ PROC_LOCK(p);
+ p-p_xstat = shared_xstat;
+ PROC_UNLOCK(p);
+ }
Why is process lock taken there ? The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
+
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
int null = 0;
@@ -257,6 +265,9 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
+ if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ /* Kill threads regardless of imgp-sysent value */
+ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
This is better expressed by
if ((p-p_sysent-sv_flags SV_ABI_MASK) == SV_ABI_LINUX)
Regardless of this mostly cosmetic issue, this is racy. Other
linux thread in the same process might do an execve(3).
More, if execve(3) call fails, then you return into the process
that lacks all threads except the one that called execve(3).
execve(3) races in linuxulator are known, and that's not the only case:
http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/142082
But fixing it is not easy, as you've noted in other email current hook
mechanism is not good enough for it.
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215672 - head/sbin/geom/class/part
Author: ae
Date: Mon Nov 22 11:24:11 2010
New Revision: 215672
URL: http://svn.freebsd.org/changeset/base/215672
Log:
Add SIGINT handler to `gpart restore` action.
MFC after:1 week
Modified:
head/sbin/geom/class/part/geom_part.c
Modified: head/sbin/geom/class/part/geom_part.c
==
--- head/sbin/geom/class/part/geom_part.c Mon Nov 22 10:08:33 2010
(r215671)
+++ head/sbin/geom/class/part/geom_part.c Mon Nov 22 11:24:11 2010
(r215672)
@@ -38,6 +38,7 @@ __FBSDID($FreeBSD$);
#include libgeom.h
#include libutil.h
#include paths.h
+#include signal.h
#include stdint.h
#include stdio.h
#include stdlib.h
@@ -61,6 +62,7 @@ uint32_t PUBSYM(version) = 0;
static char sstart[32];
static char ssize[32];
+volatile sig_atomic_t undo_restore;
#defineGPART_AUTOFILL *
#defineGPART_FLAGS C
@@ -757,12 +759,19 @@ skip_line(const char *p)
}
static void
+gpart_sighndl(int sig __unused)
+{
+ undo_restore = 1;
+}
+
+static void
gpart_restore(struct gctl_req *req, unsigned int fl __unused)
{
struct gmesh mesh;
struct gclass *classp;
struct gctl_req *r;
struct ggeom *gp;
+ struct sigaction si_sa;
const char *s, *flags, *errstr, *label;
char **ap, *argv[6], line[BUFSIZ], *pline;
int error, forced, i, l, nargs, created, rl;
@@ -786,6 +795,13 @@ gpart_restore(struct gctl_req *req, unsi
geom_deletetree(mesh);
errx(EXIT_FAILURE, Class %s not found., s);
}
+
+ sigemptyset(si_sa.sa_mask);
+ si_sa.sa_flags = 0;
+ si_sa.sa_handler = gpart_sighndl;
+ if (sigaction(SIGINT, si_sa, 0) == -1)
+ err(EXIT_FAILURE, sigaction SIGINT);
+
if (forced) {
/* destroy existent partition table before restore */
for (i = 0; i nargs; i++) {
@@ -811,7 +827,8 @@ gpart_restore(struct gctl_req *req, unsi
}
}
created = 0;
- while (fgets(line, sizeof(line) - 1, stdin)) {
+ while (undo_restore == 0
+ fgets(line, sizeof(line) - 1, stdin) != NULL) {
/* Format of backup entries:
* scheme name number of entries
* index type start size [label] ['['attrib[,attrib]']']
@@ -920,6 +937,8 @@ gpart_restore(struct gctl_req *req, unsi
}
}
}
+ if (undo_restore)
+ goto backout;
/* commit changes if needed */
if (strchr(flags, 'C') != NULL) {
for (i = 0; i nargs; i++) {
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215673 - head/sys/netgraph
Author: zec Date: Mon Nov 22 12:32:19 2010 New Revision: 215673 URL: http://svn.freebsd.org/changeset/base/215673 Log: Allow for MTU sizes of up to ETHER_MAX_LEN_JUMBO (i.e. 9018) bytes to be configured on ng_eiface ifnets. The default MTU remains unchanged at 1500 bytes. Mark ng_eiface ifnets as IFCAP_VLAN_MTU capable, so that the associated vlan(4) ifnets may use full-sized Ethernet MTUs (1500 bytes). MFC after:3 days Modified: head/sys/netgraph/ng_eiface.c head/sys/netgraph/ng_eiface.h Modified: head/sys/netgraph/ng_eiface.c == --- head/sys/netgraph/ng_eiface.c Mon Nov 22 11:24:11 2010 (r215672) +++ head/sys/netgraph/ng_eiface.c Mon Nov 22 12:32:19 2010 (r215673) @@ -371,6 +371,8 @@ ng_eiface_constructor(node_p node) ifp-if_ioctl = ng_eiface_ioctl; ifp-if_snd.ifq_maxlen = ifqmaxlen; ifp-if_flags = (IFF_SIMPLEX | IFF_BROADCAST | IFF_MULTICAST); + ifp-if_capabilities = IFCAP_VLAN_MTU | IFCAP_JUMBO_MTU; + ifp-if_capenable = IFCAP_VLAN_MTU | IFCAP_JUMBO_MTU; /* Give this node the same name as the interface (if possible) */ if (ng_name_node(node, ifp-if_xname) != 0) Modified: head/sys/netgraph/ng_eiface.h == --- head/sys/netgraph/ng_eiface.h Mon Nov 22 11:24:11 2010 (r215672) +++ head/sys/netgraph/ng_eiface.h Mon Nov 22 12:32:19 2010 (r215673) @@ -46,7 +46,7 @@ /* MTU bounds */ #define NG_EIFACE_MTU_MIN 72 -#define NG_EIFACE_MTU_MAX 2312 +#define NG_EIFACE_MTU_MAX ETHER_MAX_LEN_JUMBO #define NG_EIFACE_MTU_DEFAULT 1500 /* Netgraph commands */ ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215674 - head/sbin/devd
Author: brucec Date: Mon Nov 22 12:33:48 2010 New Revision: 215674 URL: http://svn.freebsd.org/changeset/base/215674 Log: Don't generate input() since it's not used. Modified: head/sbin/devd/token.l Modified: head/sbin/devd/token.l == --- head/sbin/devd/token.l Mon Nov 22 12:32:19 2010(r215673) +++ head/sbin/devd/token.l Mon Nov 22 12:33:48 2010(r215674) @@ -38,6 +38,7 @@ int lineno = 1; #define YY_NO_UNPUT +#define YY_NO_INPUT static void update_lineno(const char *cp) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215675 - head/sys/compat/linux
Author: netchild
Date: Mon Nov 22 12:42:32 2010
New Revision: 215675
URL: http://svn.freebsd.org/changeset/base/215675
Log:
Do not take the process lock. The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
Use a more appropriate conditional to detect the linux ABI.
Suggested by: kib
X-MFC:together with r215664
Modified:
head/sys/compat/linux/linux_emul.c
Modified: head/sys/compat/linux/linux_emul.c
==
--- head/sys/compat/linux/linux_emul.c Mon Nov 22 12:33:48 2010
(r215674)
+++ head/sys/compat/linux/linux_emul.c Mon Nov 22 12:42:32 2010
(r215675)
@@ -198,11 +198,8 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
- if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
- PROC_LOCK(p);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0)
p-p_xstat = shared_xstat;
- PROC_UNLOCK(p);
- }
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
@@ -265,7 +262,8 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
- if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) ==
+ SV_ABI_LINUX))
/* Kill threads regardless of imgp-sysent value */
linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
if (__predict_false(imgp-sysent != elf_linux_sysvec
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215675 - head/sys/compat/linux
On Mon, Nov 22, 2010 at 12:42:32PM +, Alexander Leidinger wrote: Author: netchild Date: Mon Nov 22 12:42:32 2010 New Revision: 215675 URL: http://svn.freebsd.org/changeset/base/215675 Log: Do not take the process lock. The assignment to u_short inside the properly aligned structure is atomic on all supported architectures, and the thread that should see side-effect of assignment is the same thread that does assignment. Use a more appropriate conditional to detect the linux ABI. Suggested by: kib X-MFC: together with r215664 Modified: head/sys/compat/linux/linux_emul.c Modified: head/sys/compat/linux/linux_emul.c == --- head/sys/compat/linux/linux_emul.cMon Nov 22 12:33:48 2010 (r215674) +++ head/sys/compat/linux/linux_emul.cMon Nov 22 12:42:32 2010 (r215675) @@ -265,7 +262,8 @@ linux_proc_exec(void *arg __unused, stru if (__predict_false(imgp-sysent == elf_linux_sysvec p-p_sysent != elf_linux_sysvec)) linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0); - if (__predict_false(p-p_sysent == elf_linux_sysvec)) + if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) == + SV_ABI_LINUX)) /* Kill threads regardless of imgp-sysent value */ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL); if (__predict_false(imgp-sysent != elf_linux_sysvec There are several similar comparisons around the patched one. I am still quite curious for the reason of all __predict() obfuscations that are countless in the linuxolator. We are not oblidged to emulate this aspect of Linux. pgpIeifMMWQZN.pgp Description: PGP signature
svn commit: r215676 - in head: sbin/hastd usr.bin/colldef usr.sbin/apmd usr.sbin/bluetooth/bthidd usr.sbin/bluetooth/hcsecd usr.sbin/config usr.sbin/kbdcontrol
Author: brucec
Date: Mon Nov 22 14:16:22 2010
New Revision: 215676
URL: http://svn.freebsd.org/changeset/base/215676
Log:
Don't generate input() since it's not used.
Modified:
head/sbin/hastd/Makefile
head/usr.bin/colldef/Makefile
head/usr.sbin/apmd/apmdlex.l
head/usr.sbin/bluetooth/bthidd/lexer.l
head/usr.sbin/bluetooth/hcsecd/lexer.l
head/usr.sbin/config/lang.l
head/usr.sbin/kbdcontrol/lex.l
Modified: head/sbin/hastd/Makefile
==
--- head/sbin/hastd/MakefileMon Nov 22 12:42:32 2010(r215675)
+++ head/sbin/hastd/MakefileMon Nov 22 14:16:22 2010(r215676)
@@ -25,6 +25,7 @@ CFLAGS+=-DINET6
.endif
# This is needed to have WARNS 1.
CFLAGS+=-DYY_NO_UNPUT
+CFLAGS+=-DYY_NO_INPUT
DPADD= ${LIBGEOM} ${LIBBSDXML} ${LIBSBUF} ${LIBL} ${LIBPTHREAD} ${LIBUTIL}
LDADD= -lgeom -lbsdxml -lsbuf -ll -lpthread -lutil
Modified: head/usr.bin/colldef/Makefile
==
--- head/usr.bin/colldef/Makefile Mon Nov 22 12:42:32 2010
(r215675)
+++ head/usr.bin/colldef/Makefile Mon Nov 22 14:16:22 2010
(r215676)
@@ -4,7 +4,7 @@ PROG= colldef
SRCS= parse.y scan.l y.tab.h
LFLAGS=-8 -i
CFLAGS+=-I. -I${.CURDIR} -I${.CURDIR}/../../lib/libc/locale
-CFLAGS+=-DCOLLATE_DEBUG -DYY_NO_UNPUT
+CFLAGS+=-DCOLLATE_DEBUG -DYY_NO_UNPUT -DYY_NO_INPUT
LDADD= -ll
DPADD= ${LIBL}
Modified: head/usr.sbin/apmd/apmdlex.l
==
--- head/usr.sbin/apmd/apmdlex.lMon Nov 22 12:42:32 2010
(r215675)
+++ head/usr.sbin/apmd/apmdlex.lMon Nov 22 14:16:22 2010
(r215676)
@@ -38,6 +38,7 @@
/* We don't need it, avoid the warning. */
#define YY_NO_UNPUT
+#define YY_NO_INPUT
int lineno;
int first_time;
Modified: head/usr.sbin/bluetooth/bthidd/lexer.l
==
--- head/usr.sbin/bluetooth/bthidd/lexer.l Mon Nov 22 12:42:32 2010
(r215675)
+++ head/usr.sbin/bluetooth/bthidd/lexer.l Mon Nov 22 14:16:22 2010
(r215676)
@@ -39,7 +39,7 @@
int yylex (void);
%}
-%option yylineno noyywrap nounput
+%option yylineno noyywrap nounput noinput
delim [ \t\n]
ws {delim}+
Modified: head/usr.sbin/bluetooth/hcsecd/lexer.l
==
--- head/usr.sbin/bluetooth/hcsecd/lexer.l Mon Nov 22 12:42:32 2010
(r215675)
+++ head/usr.sbin/bluetooth/hcsecd/lexer.l Mon Nov 22 14:16:22 2010
(r215676)
@@ -34,7 +34,7 @@
#include parser.h
%}
-%option yylineno noyywrap nounput
+%option yylineno noyywrap nounput noinput
delim [ \t\n]
ws {delim}+
Modified: head/usr.sbin/config/lang.l
==
--- head/usr.sbin/config/lang.l Mon Nov 22 12:42:32 2010(r215675)
+++ head/usr.sbin/config/lang.l Mon Nov 22 14:16:22 2010(r215676)
@@ -39,6 +39,7 @@
#include config.h
#define YY_NO_UNPUT
+#define YY_NO_INPUT
/*
* Data for returning to previous files from include files.
Modified: head/usr.sbin/kbdcontrol/lex.l
==
--- head/usr.sbin/kbdcontrol/lex.l Mon Nov 22 12:42:32 2010
(r215675)
+++ head/usr.sbin/kbdcontrol/lex.l Mon Nov 22 14:16:22 2010
(r215676)
@@ -32,6 +32,7 @@
#include lex.h
#define YY_NO_UNPUT
+#define YY_NO_INPUT
%}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215677 - head/sys/netinet
Author: zec Date: Mon Nov 22 14:16:23 2010 New Revision: 215677 URL: http://svn.freebsd.org/changeset/base/215677 Log: Remove an apparently redundant CURVNET_SET() / CURVNET_RESTORE() pair. MFC after:3 days Modified: head/sys/netinet/if_ether.c Modified: head/sys/netinet/if_ether.c == --- head/sys/netinet/if_ether.c Mon Nov 22 14:16:22 2010(r215676) +++ head/sys/netinet/if_ether.c Mon Nov 22 14:16:23 2010(r215677) @@ -154,12 +154,10 @@ arp_ifscrub(struct ifnet *ifp, uint32_t addr4.sin_len= sizeof(addr4); addr4.sin_family = AF_INET; addr4.sin_addr.s_addr = addr; - CURVNET_SET(ifp-if_vnet); IF_AFDATA_LOCK(ifp); lla_lookup(LLTABLE(ifp), (LLE_DELETE | LLE_IFADDR), (struct sockaddr *)addr4); IF_AFDATA_UNLOCK(ifp); - CURVNET_RESTORE(); } #endif ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215678 - head/usr.sbin/mptutil
Author: jhb
Date: Mon Nov 22 14:36:04 2010
New Revision: 215678
URL: http://svn.freebsd.org/changeset/base/215678
Log:
Similar to mfiutil, drop local definition of powerof2() and use version
from sys/param.h instead.
Modified:
head/usr.sbin/mptutil/mpt_config.c
Modified: head/usr.sbin/mptutil/mpt_config.c
==
--- head/usr.sbin/mptutil/mpt_config.c Mon Nov 22 14:16:23 2010
(r215677)
+++ head/usr.sbin/mptutil/mpt_config.c Mon Nov 22 14:36:04 2010
(r215678)
@@ -50,8 +50,6 @@ __RCSID($FreeBSD$);
static voiddump_config(CONFIG_PAGE_RAID_VOL_0 *vol);
#endif
-#define powerof2(x)x)-1)(x))==0)
-
static long
dehumanize(const char *value)
{
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215679 - in head: contrib/binutils/bfd contrib/binutils/binutils contrib/binutils/include/elf contrib/file contrib/gdb/gdb gnu/usr.bin/gdb/libgdb sys/compat/freebsd32 sys/kern sys/sys ...
Author: attilio
Date: Mon Nov 22 14:42:13 2010
New Revision: 215679
URL: http://svn.freebsd.org/changeset/base/215679
Log:
Add the ability for GDB to printout the thread name along with other
thread specific informations.
In order to do that, and in order to avoid KBI breakage with existing
infrastructure the following semantic is implemented:
- For live programs, a new member to the PT_LWPINFO is added (pl_tdname)
- For cores, a new ELF note is added (NT_THRMISC) that can be used for
storing thread specific, miscellaneous, informations. Right now it is
just popluated with a thread name.
GDB, then, retrieves the correct informations from the corefile via the
BFD interface, as it groks the ELF notes and create appropriate
pseudo-sections.
Sponsored by: Sandvine Incorporated
Tested by:gianni
Discussed with: dim, kan, kib
MFC after:2 weeks
Modified:
head/contrib/binutils/bfd/elf-bfd.h
head/contrib/binutils/bfd/elf.c
head/contrib/binutils/binutils/readelf.c
head/contrib/binutils/include/elf/common.h
head/contrib/file/readelf.h
head/contrib/gdb/gdb/fbsd-proc.c
head/gnu/usr.bin/gdb/libgdb/fbsd-threads.c
head/sys/compat/freebsd32/freebsd32.h
head/sys/kern/imgact_elf.c
head/sys/kern/sys_process.c
head/sys/sys/elf_common.h
head/sys/sys/procfs.h
head/sys/sys/ptrace.h
head/usr.bin/gcore/elfcore.c
Modified: head/contrib/binutils/bfd/elf-bfd.h
==
--- head/contrib/binutils/bfd/elf-bfd.h Mon Nov 22 14:36:04 2010
(r215678)
+++ head/contrib/binutils/bfd/elf-bfd.h Mon Nov 22 14:42:13 2010
(r215679)
@@ -1673,6 +1673,8 @@ extern char * elfcore_write_pstatus
(bfd *, char *, int *, long, int, const void *);
extern char *elfcore_write_prfpreg
(bfd *, char *, int *, const void *, int);
+extern char *elfcore_write_thrmisc
+ (bfd *, char *, int *, const char *, int);
extern char *elfcore_write_prxfpreg
(bfd *, char *, int *, const void *, int);
extern char *elfcore_write_lwpstatus
Modified: head/contrib/binutils/bfd/elf.c
==
--- head/contrib/binutils/bfd/elf.c Mon Nov 22 14:36:04 2010
(r215678)
+++ head/contrib/binutils/bfd/elf.c Mon Nov 22 14:42:13 2010
(r215679)
@@ -6316,6 +6316,12 @@ _bfd_elf_rel_vtable_reloc_fn
#ifdef HAVE_SYS_PROCFS_H
# include sys/procfs.h
+
+/* Define HAVE_THRMISC_T for consistency with other similar GNU-type stubs. */
+#undef HAVE_THRMISC_T
+#if defined (THRMISC_VERSION)
+#defineHAVE_THRMISC_T 1
+#endif
#endif
/* FIXME: this is kinda wrong, but it's what gdb wants. */
@@ -6497,6 +6503,16 @@ elfcore_grok_prxfpreg (bfd *abfd, Elf_In
return elfcore_make_note_pseudosection (abfd, .reg-xfp, note);
}
+#if defined (HAVE_THRMISC_T)
+
+static bfd_boolean
+elfcore_grok_thrmisc (bfd *abfd, Elf_Internal_Note *note)
+{
+ return elfcore_make_note_pseudosection (abfd, .tname, note);
+}
+
+#endif /* defined (HAVE_THRMISC_T) */
+
#if defined (HAVE_PRPSINFO_T)
typedef prpsinfo_t elfcore_psinfo_t;
#if defined (HAVE_PRPSINFO32_T)/* Sparc64 cross Sparc32 */
@@ -6863,6 +6879,12 @@ elfcore_grok_note (bfd *abfd, Elf_Intern
return TRUE;
}
+
+#if defined (HAVE_THRMISC_T)
+case NT_THRMISC:
+ return elfcore_grok_thrmisc (abfd, note);
+#endif
+
}
}
@@ -7245,6 +7267,22 @@ elfcore_write_prfpreg (bfd *abfd,
}
char *
+elfcore_write_thrmisc (bfd *abfd,
+ char *buf,
+ int *bufsiz,
+ const char *tname,
+ int size)
+{
+#if defined (HAVE_THRMISC_T)
+ char *note_name = CORE;
+ return elfcore_write_note (abfd, buf, bufsiz,
+note_name, NT_THRMISC, tname, size);
+#else
+ return buf;
+#endif
+}
+
+char *
elfcore_write_prxfpreg (bfd *abfd,
char *buf,
int *bufsiz,
Modified: head/contrib/binutils/binutils/readelf.c
==
--- head/contrib/binutils/binutils/readelf.cMon Nov 22 14:36:04 2010
(r215678)
+++ head/contrib/binutils/binutils/readelf.cMon Nov 22 14:42:13 2010
(r215679)
@@ -9908,6 +9908,7 @@ get_note_type (unsigned e_type)
case NT_PSTATUS: return _(NT_PSTATUS (pstatus structure));
case NT_FPREGS:return _(NT_FPREGS (floating point registers));
case NT_PSINFO:return _(NT_PSINFO (psinfo structure));
+case NT_THRMISC: return _(NT_THRMISC (thrmisc structure));
case NT_LWPSTATUS: return _(NT_LWPSTATUS (lwpstatus_t structure));
case NT_LWPSINFO: return _(NT_LWPSINFO (lwpsinfo_t structure));
case NT_WIN32PSTATUS: return _(NT_WIN32PSTATUS (win32_pstatus
structure));
Modified: head/contrib/binutils/include/elf/common.h
svn commit: r215680 - head/lib/libpam/modules/pam_unix
Author: des Date: Mon Nov 22 14:45:16 2010 New Revision: 215680 URL: http://svn.freebsd.org/changeset/base/215680 Log: Add time.h for ctime(), which we accidentally picked up through sys/time.h. Submitted by: Garrett Cooper yaneg...@gmail.com MFC after:3 days Modified: head/lib/libpam/modules/pam_unix/pam_unix.c Modified: head/lib/libpam/modules/pam_unix/pam_unix.c == --- head/lib/libpam/modules/pam_unix/pam_unix.c Mon Nov 22 14:42:13 2010 (r215679) +++ head/lib/libpam/modules/pam_unix/pam_unix.c Mon Nov 22 14:45:16 2010 (r215680) @@ -50,6 +50,7 @@ __FBSDID($FreeBSD$); #include string.h #include stdio.h #include syslog.h +#include time.h #include unistd.h #include libutil.h ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215544 - head/sys/kern
On Friday, November 19, 2010 5:41:00 pm Attilio Rao wrote:
2010/11/19 John Baldwin j...@freebsd.org:
On Friday, November 19, 2010 5:03:25 pm Jung-uk Kim wrote:
On Friday 19 November 2010 04:46 pm, John Baldwin wrote:
On Friday, November 19, 2010 4:31:44 pm Attilio Rao wrote:
2010/11/19 John Baldwin j...@freebsd.org:
On Friday, November 19, 2010 4:09:28 pm Jung-uk Kim wrote:
On Friday 19 November 2010 02:43 pm, Attilio Rao wrote:
Author: attilio
Date: Fri Nov 19 19:43:56 2010
New Revision: 215544
URL: http://svn.freebsd.org/changeset/base/215544
Log:
Scan the list in reverse order for the shutdown handlers
of loaded modules. This way, when there is a dependency
between two modules, the handler of the latter probed runs
first.
This is a similar approach as the modules are unloaded in
the same linkerfile.
Sponsored by: Sandvine Incorporated
Submitted by: Nima Misaghian nmisaghian at sandvine
dot com MFC after:1 week
Hmm... It is not directly related but I was thinking about
doing similar things for sys/kern/subr_bus.c. What do you
think about the attached patch?
Hmm, the patches for suspend and resume that I had for this
took the opposite order, they did suspend in forward order, but
resume in backwards order. Like so:
--- //depot/vendor/freebsd/src/sys/kern/subr_bus.c
2010-11-17 22:30:24.0 +++
//depot/user/jhb/acpipci/kern/subr_bus.c2010-11-19
17:19:02.0 00 @@ -3426,9 +3429,9 @@
TAILQ_FOREACH(child, dev-children, link) {
error = DEVICE_SUSPEND(child);
if (error) {
- for (child2 =
TAILQ_FIRST(dev-children); -
child2 child2 != child; -child2
= TAILQ_NEXT(child2, link)) + for (child2
= TAILQ_PREV(child, device_list, link); +
child2 != NULL;
+child2 = TAILQ_PREV(child2,
device_list, link)) DEVICE_RESUME(child2);
return (error);
}
@@ -3447,7 +3450,7 @@
{
device_tchild;
- TAILQ_FOREACH(child, dev-children, link) {
+ TAILQ_FOREACH_REVERSE(child, dev-children,
device_list, link) { DEVICE_RESUME(child);
/* if resume fails, there's nothing we can
usefully do... */ }
(Likely mangled whitespace.)
I couldn't convince myself which order was more correct for
suspend and resume.
Considering loading in starting point, I think suspend should go
in reverse logic and resume in the same module load logic.
So that dependent modules are suspended first and resumed after.
Don't you agree?
These are devices, and the ordering here is the order of sibling
devices on a given bus. That is, if you have a PCI bus with two em
interfaces, does it really matter if em0 suspends before em1 vs
after em1? I think it actually doesn't matter. The passes from
the multipass boot probe might make some sense to order on.
However, I don't think the order of siblings on a bus is meaningful
for suspend and resume (which is why I've never committed the above
patches).
Specifically, there is no dependency relationship between siblings
on a bus. Certain buses may in fact have a dependency order of
sorts (vgapci0 comes to mind), but those buses should manage that.
There is no generic dependency between siblings that should be
encoded into subr_bus.c
Generally siblings don't interact with each other directly, no.
However, some modern chipsets *do* have strong relationship. For
example, some chipsets reference SMB controller to get current
configuration, e.g., function A depends on function B on the same
chip.
That may be true, but that is not generic to all buses and devices.
That isn't even really generic to PCI. If there are specific instances
where there are dependencies, the drivers for that hardware should
manage that. If specific buses have specific orders, then they can
manage that order explicitly in their own suspend and resume routines.
However, I don't see a valid reason for enforcing a particular order
among siblings for all devices. We certainly do enforce some orders
with respect to children and parents (parents attach first and resume
first, children detach first and suspend first, PCI even mandates this
for powering down a bus), but the same is not true for siblings.
However, it would be good if any ordering is maintained between
devices linked via MODDEPEND stuff.
It happens they may also be sibillings, and not havign parent-child
relationship in some cases, but I'm also wondering if maybe they
should ensure consistency in their code itself
Re: svn commit: r215544 - head/sys/kern
On Friday, November 19, 2010 5:32:00 pm Attilio Rao wrote: 2010/11/19 John Baldwin j...@freebsd.org: On Friday, November 19, 2010 4:27:27 pm Attilio Rao wrote: 2010/11/19 John Baldwin j...@freebsd.org: On Friday, November 19, 2010 2:43:57 pm Attilio Rao wrote: Author: attilio Date: Fri Nov 19 19:43:56 2010 New Revision: 215544 URL: http://svn.freebsd.org/changeset/base/215544 Log: Scan the list in reverse order for the shutdown handlers of loaded modules. This way, when there is a dependency between two modules, the handler of the latter probed runs first. This is a similar approach as the modules are unloaded in the same linkerfile. Sponsored by: Sandvine Incorporated Submitted by: Nima Misaghian nmisaghian at sandvine dot com MFC after: 1 week Modified: head/sys/kern/kern_module.c Modified: head/sys/kern/kern_module.c == --- head/sys/kern/kern_module.c Fri Nov 19 18:59:35 2010 (r215543) +++ head/sys/kern/kern_module.c Fri Nov 19 19:43:56 2010 (r215544) @@ -46,7 +46,7 @@ __FBSDID($FreeBSD$); static MALLOC_DEFINE(M_MODULE, module, module data structures); -typedef TAILQ_HEAD(, module) modulelist_t; +typedef TAILQ_HEAD(modulelst, module) modulelist_t; Is modulelist already taken? If not, we should probably just retire 'modulelist_t' and replace it with 'struct modulelist'. Note that I used modulelst, not modulelist. Probabilly, if you think the name may be still confusing, we can pickup one another further. Yes, I'd much prefer 'modulelist' with the extra 'i' as it is more readable. If you go that route, I think you can drop modulelist_t since style(9) discourages foo_t types unless they are required by a standard. I think that this patch should make it? Works for me, thanks. -- John Baldwin ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215681 - in head/sys: . amd64/include/xen cddl/contrib/opensolaris contrib/x86emu dev/xen/xenpci
Author: jhb Date: Mon Nov 22 15:15:11 2010 New Revision: 215681 URL: http://svn.freebsd.org/changeset/base/215681 Log: Remove some bogus, self-referential mergeinfo. Modified: Directory Properties: head/sys/ (props changed) head/sys/amd64/include/xen/ (props changed) head/sys/cddl/contrib/opensolaris/ (props changed) head/sys/contrib/x86emu/ (props changed) head/sys/dev/xen/xenpci/ (props changed) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215682 - head/sys/dev/xen/xenpci
Author: jhb Date: Mon Nov 22 15:26:47 2010 New Revision: 215682 URL: http://svn.freebsd.org/changeset/base/215682 Log: Purge mergeinfo on sys/dev/xen/xenpci. The only unique mergeinfo compared to head was not useful (it came in with the merge from /user/dfr/xenhvm/7 and that mergeinfo is still present at sys/) and not worth keeping an extra set of mergeinfo around in the kernel. Modified: Directory Properties: head/sys/dev/xen/xenpci/ (props changed) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215683 - head/sys/kern
Author: attilio
Date: Mon Nov 22 15:28:54 2010
New Revision: 215683
URL: http://svn.freebsd.org/changeset/base/215683
Log:
Style fix.
Sponsored by: Sandvine Incorporated
Requested by: jhb
Reviewed by: jhb
MFC after:1 week
X-MFC:215544
Modified:
head/sys/kern/kern_module.c
Modified: head/sys/kern/kern_module.c
==
--- head/sys/kern/kern_module.c Mon Nov 22 15:26:47 2010(r215682)
+++ head/sys/kern/kern_module.c Mon Nov 22 15:28:54 2010(r215683)
@@ -46,7 +46,6 @@ __FBSDID($FreeBSD$);
static MALLOC_DEFINE(M_MODULE, module, module data structures);
-typedef TAILQ_HEAD(modulelst, module) modulelist_t;
struct module {
TAILQ_ENTRY(module) link; /* chain together all modules */
TAILQ_ENTRY(module) flink; /* all modules in a file */
@@ -61,7 +60,7 @@ struct module {
#define MOD_EVENT(mod, type) (mod)-handler((mod), (type), (mod)-arg)
-static modulelist_t modules;
+static TAILQ_HEAD(modulelist, module) modules;
struct sx modules_sx;
static int nextid = 1;
static void module_shutdown(void *, int);
@@ -101,7 +100,7 @@ module_shutdown(void *arg1, int arg2)
return;
mtx_lock(Giant);
MOD_SLOCK;
- TAILQ_FOREACH_REVERSE(mod, modules, modulelst, link)
+ TAILQ_FOREACH_REVERSE(mod, modules, modulelist, link)
MOD_EVENT(mod, MOD_SHUTDOWN);
MOD_SUNLOCK;
mtx_unlock(Giant);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215675 - head/sys/compat/linux
On Monday, November 22, 2010 7:51:06 am Kostik Belousov wrote: On Mon, Nov 22, 2010 at 12:42:32PM +, Alexander Leidinger wrote: Author: netchild Date: Mon Nov 22 12:42:32 2010 New Revision: 215675 URL: http://svn.freebsd.org/changeset/base/215675 Log: Do not take the process lock. The assignment to u_short inside the properly aligned structure is atomic on all supported architectures, and the thread that should see side-effect of assignment is the same thread that does assignment. Use a more appropriate conditional to detect the linux ABI. Suggested by: kib X-MFC:together with r215664 Modified: head/sys/compat/linux/linux_emul.c Modified: head/sys/compat/linux/linux_emul.c == --- head/sys/compat/linux/linux_emul.c Mon Nov 22 12:33:48 2010 (r215674) +++ head/sys/compat/linux/linux_emul.c Mon Nov 22 12:42:32 2010 (r215675) @@ -265,7 +262,8 @@ linux_proc_exec(void *arg __unused, stru if (__predict_false(imgp-sysent == elf_linux_sysvec p-p_sysent != elf_linux_sysvec)) linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0); - if (__predict_false(p-p_sysent == elf_linux_sysvec)) + if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) == + SV_ABI_LINUX)) /* Kill threads regardless of imgp-sysent value */ linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL); if (__predict_false(imgp-sysent != elf_linux_sysvec There are several similar comparisons around the patched one. I am still quite curious for the reason of all __predict() obfuscations that are countless in the linuxolator. We are not oblidged to emulate this aspect of Linux. I think we should only have __predict() obfuscation if there is a benchmark showing an indisputable performance gain. Otherwise it is just cruft making the code hard to read. I suspect that a Linux app calling execve() is not a sufficient critical path to warrant any of these. -- John Baldwin ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215684 - stable/8/sys/conf
Author: kensmith
Date: Mon Nov 22 16:09:57 2010
New Revision: 215684
URL: http://svn.freebsd.org/changeset/base/215684
Log:
We're a bit under a week from Code Freeze for the upcoming 8.2-RELEASE
cycle. Warn people tracking stable/8 that the branch may be more
active than usual.
Modified:
stable/8/sys/conf/newvers.sh
Modified: stable/8/sys/conf/newvers.sh
==
--- stable/8/sys/conf/newvers.shMon Nov 22 15:28:54 2010
(r215683)
+++ stable/8/sys/conf/newvers.shMon Nov 22 16:09:57 2010
(r215684)
@@ -31,8 +31,8 @@
# $FreeBSD$
TYPE=FreeBSD
-REVISION=8.1
-BRANCH=STABLE
+REVISION=8.2
+BRANCH=PRERELEASE
if [ X${BRANCH_OVERRIDE} != X ]; then
BRANCH=${BRANCH_OVERRIDE}
fi
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215685 - stable/7/sys/conf
Author: kensmith
Date: Mon Nov 22 16:10:54 2010
New Revision: 215685
URL: http://svn.freebsd.org/changeset/base/215685
Log:
We're a bit under a week from Code Freeze for the upcoming 7.4-RELEASE
cycle. Warn people tracking stable/7 that the branch may be more
active than usual.
Modified:
stable/7/sys/conf/newvers.sh
Modified: stable/7/sys/conf/newvers.sh
==
--- stable/7/sys/conf/newvers.shMon Nov 22 16:09:57 2010
(r215684)
+++ stable/7/sys/conf/newvers.shMon Nov 22 16:10:54 2010
(r215685)
@@ -31,8 +31,8 @@
# $FreeBSD$
TYPE=FreeBSD
-REVISION=7.3
-BRANCH=STABLE
+REVISION=7.4
+BRANCH=PRERELEASE
if [ X${BRANCH_OVERRIDE} != X ]; then
BRANCH=${BRANCH_OVERRIDE}
fi
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215686 - head/sys/dev/mxge
Author: gallatin
Date: Mon Nov 22 16:43:05 2010
New Revision: 215686
URL: http://svn.freebsd.org/changeset/base/215686
Log:
Fix a TSO checksum bug on mxge(4):
The Myri10GE NIC will assume all TSO frames contain partial checksum,
and will emit TSO segments with bad TCP checksums if a TSO frame
contains a full checksum. The mxge driver takes care to make sure
that TSO is disabled when checksum offload is disabled for this
reason. However, modules that modify packet contents (like pf) may
end up completing a checksum on a TSO frame, leading to the NIC emitting
TSO segments with bad checksums.
To workaround this, restore the partial checksum in the mxge driver
when we're fed a TSO frame with a full checksum.
Reported by: Bob Healey
MFC after:3 days
Modified:
head/sys/dev/mxge/if_mxge.c
Modified: head/sys/dev/mxge/if_mxge.c
==
--- head/sys/dev/mxge/if_mxge.c Mon Nov 22 16:10:54 2010(r215685)
+++ head/sys/dev/mxge/if_mxge.c Mon Nov 22 16:43:05 2010(r215686)
@@ -1855,9 +1855,20 @@ mxge_encap_tso(struct mxge_slice_state *
tcp = (struct tcphdr *)((char *)ip + (ip-ip_hl 2));
cum_len = -(ip_off + ((ip-ip_hl + tcp-th_off) 2));
+ cksum_offset = ip_off + (ip-ip_hl 2);
/* TSO implies checksum offload on this hardware */
- cksum_offset = ip_off + (ip-ip_hl 2);
+ if (__predict_false((m-m_pkthdr.csum_flags (CSUM_TCP)) == 0)) {
+ /*
+* If packet has full TCP csum, replace it with pseudo hdr
+* sum that the NIC expects, otherwise the NIC will emit
+* packets with bad TCP checksums.
+*/
+ m-m_pkthdr.csum_flags = CSUM_TCP;
+ m-m_pkthdr.csum_data = offsetof(struct tcphdr, th_sum);
+ tcp-th_sum = in_pseudo(ip-ip_src.s_addr, ip-ip_dst.s_addr,
+ htons(IPPROTO_TCP + (m-m_pkthdr.len - cksum_offset)));
+ }
flags = MXGEFW_FLAGS_TSO_HDR | MXGEFW_FLAGS_FIRST;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215687 - head/sys/geom
Author: jh
Date: Mon Nov 22 16:47:53 2010
New Revision: 215687
URL: http://svn.freebsd.org/changeset/base/215687
Log:
Use g_eventlock to protect against losing wakeups in the g_event process
and replace tsleep(9) with msleep(9) which doesn't use a timeout. The
previously used timeout caused the event process to wake up ten times
per second on an idle system.
one_event() is now called with the topology lock held and it returns
with both the topology and event locks held when there are no more
events in the queue.
Reported by: mav, Marius Nünnerich
Reviewed by: freebsd-geom
Modified:
head/sys/geom/geom_event.c
head/sys/geom/geom_kern.c
Modified: head/sys/geom/geom_event.c
==
--- head/sys/geom/geom_event.c Mon Nov 22 16:43:05 2010(r215686)
+++ head/sys/geom/geom_event.c Mon Nov 22 16:47:53 2010(r215687)
@@ -183,33 +183,27 @@ one_event(void)
struct g_event *ep;
struct g_provider *pp;
- g_topology_lock();
- for (;;) {
- mtx_lock(g_eventlock);
- TAILQ_FOREACH(pp, g_doorstep, orphan) {
- if (pp-nstart == pp-nend)
- break;
- }
- if (pp != NULL) {
- G_VALID_PROVIDER(pp);
- TAILQ_REMOVE(g_doorstep, pp, orphan);
- }
- mtx_unlock(g_eventlock);
- if (pp == NULL)
+ g_topology_assert();
+ mtx_lock(g_eventlock);
+ TAILQ_FOREACH(pp, g_doorstep, orphan) {
+ if (pp-nstart == pp-nend)
break;
+ }
+ if (pp != NULL) {
+ G_VALID_PROVIDER(pp);
+ TAILQ_REMOVE(g_doorstep, pp, orphan);
+ mtx_unlock(g_eventlock);
g_orphan_register(pp);
+ return (1);
}
- mtx_lock(g_eventlock);
+
ep = TAILQ_FIRST(g_events);
if (ep == NULL) {
wakeup(g_pending_events);
- mtx_unlock(g_eventlock);
- g_topology_unlock();
return (0);
}
if (ep-flag EV_INPROGRESS) {
mtx_unlock(g_eventlock);
- g_topology_unlock();
return (1);
}
ep-flag |= EV_INPROGRESS;
@@ -228,7 +222,6 @@ one_event(void)
mtx_unlock(g_eventlock);
g_free(ep);
}
- g_topology_unlock();
return (1);
}
@@ -237,16 +230,27 @@ g_run_events()
{
int i;
- while (one_event())
- ;
- g_topology_lock();
- i = g_wither_work;
- while (i) {
- i = g_wither_washer();
- g_wither_work = i 1;
- i = 2;
+ for (;;) {
+ g_topology_lock();
+ while (one_event())
+ ;
+ mtx_assert(g_eventlock, MA_OWNED);
+ i = g_wither_work;
+ if (i) {
+ mtx_unlock(g_eventlock);
+ while (i) {
+ i = g_wither_washer();
+ g_wither_work = i 1;
+ i = 2;
+ }
+ g_topology_unlock();
+ } else {
+ g_topology_unlock();
+ msleep(g_wait_event, g_eventlock, PRIBIO | PDROP,
+ -, 0);
+ }
}
- g_topology_unlock();
+ /* NOTREACHED */
}
void
@@ -338,9 +342,12 @@ g_post_event(g_event_t *func, void *arg,
}
void
-g_do_wither() {
+g_do_wither()
+{
+ mtx_lock(g_eventlock);
g_wither_work = 1;
+ mtx_unlock(g_eventlock);
wakeup(g_wait_event);
}
Modified: head/sys/geom/geom_kern.c
==
--- head/sys/geom/geom_kern.c Mon Nov 22 16:43:05 2010(r215686)
+++ head/sys/geom/geom_kern.c Mon Nov 22 16:47:53 2010(r215687)
@@ -137,10 +137,8 @@ g_event_procbody(void)
thread_lock(tp);
sched_prio(tp, PRIBIO);
thread_unlock(tp);
- for(;;) {
- g_run_events();
- tsleep(g_wait_event, PRIBIO, -, hz/10);
- }
+ g_run_events();
+ /* NOTREACHED */
}
static struct kproc_desc g_event_kp = {
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215688 - stable/8/sys/kern
Author: nwhitehorn Date: Mon Nov 22 16:58:07 2010 New Revision: 215688 URL: http://svn.freebsd.org/changeset/base/215688 Log: MFC r213986: Fix an XXX comment by answering 'no'. OS X does not set the day-of-week counter on SMU-based systems, which causes FreeBSD to reject the RTC time when used in a dual-boot environment. Since we don't use the day-of-week counter anyway, solve this by just not checking that it matches. Modified: stable/8/sys/kern/subr_clock.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/amd64/include/xen/ (props changed) stable/8/sys/cddl/contrib/opensolaris/ (props changed) stable/8/sys/contrib/dev/acpica/ (props changed) stable/8/sys/contrib/pf/ (props changed) stable/8/sys/dev/xen/xenpci/ (props changed) Modified: stable/8/sys/kern/subr_clock.c == --- stable/8/sys/kern/subr_clock.c Mon Nov 22 16:47:53 2010 (r215687) +++ stable/8/sys/kern/subr_clock.c Mon Nov 22 16:58:07 2010 (r215688) @@ -165,10 +165,6 @@ clock_ct_to_ts(struct clocktime *ct, str days += days_in_month(year, i); days += (ct-day - 1); - /* XXX Dow sanity check. Dow is not used, so should we check it? */ - if (ct-dow != -1 ct-dow != day_of_week(days)) - return (EINVAL); - /* Add hours, minutes, seconds. */ secs = ((days * 24 + ct-hour) * 60 + ct-min) * 60 + ct-sec; ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215689 - stable/8/sys/powerpc/aim
Author: nwhitehorn
Date: Mon Nov 22 17:01:40 2010
New Revision: 215689
URL: http://svn.freebsd.org/changeset/base/215689
Log:
MFC r214610:
Map userland copyin/copyout segment no-execute while it is mapped into the
kernel.
Modified:
stable/8/sys/powerpc/aim/copyinout.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/aim/copyinout.c
==
--- stable/8/sys/powerpc/aim/copyinout.cMon Nov 22 16:58:07 2010
(r215688)
+++ stable/8/sys/powerpc/aim/copyinout.cMon Nov 22 17:01:40 2010
(r215689)
@@ -76,6 +76,9 @@ static __inline void
set_user_sr(register_t vsid)
{
+ /* Mark segment no-execute */
+ vsid |= SR_N;
+
isync();
__asm __volatile (mtsr %0,%1 :: n(USER_SR), r(vsid));
isync();
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215690 - stable/8/sys/powerpc/powermac
Author: nwhitehorn
Date: Mon Nov 22 17:03:09 2010
New Revision: 215690
URL: http://svn.freebsd.org/changeset/base/215690
Log:
MFC r214575:
Allow access to the HT I/O port space on the IBM CPC9X5 northbridge chips.
Modified:
stable/8/sys/powerpc/powermac/cpcht.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/powermac/cpcht.c
==
--- stable/8/sys/powerpc/powermac/cpcht.c Mon Nov 22 17:01:40 2010
(r215689)
+++ stable/8/sys/powerpc/powermac/cpcht.c Mon Nov 22 17:03:09 2010
(r215690)
@@ -142,6 +142,7 @@ struct cpcht_softc {
vm_offset_t sc_data;
uint64_tsc_populated_slots;
struct rman sc_mem_rman;
+ struct rman sc_io_rman;
struct cpcht_irqhtirq_map[128];
};
@@ -156,6 +157,9 @@ static devclass_t cpcht_devclass;
DRIVER_MODULE(cpcht, nexus, cpcht_driver, cpcht_devclass, 0, 0);
+#define CPCHT_IOPORT_BASE 0xf400UL /* Hardwired */
+#define CPCHT_IOPORT_SIZE 0x0040UL
+
#define HTAPIC_REQUEST_EOI 0x20
#define HTAPIC_TRIGGER_LEVEL 0x02
#define HTAPIC_MASK0x01
@@ -215,7 +219,14 @@ cpcht_attach(device_t dev)
sc-sc_mem_rman.rm_type = RMAN_ARRAY;
sc-sc_mem_rman.rm_descr = CPCHT Device Memory;
error = rman_init(sc-sc_mem_rman);
+ if (error) {
+ device_printf(dev, rman_init() failed. error = %d\n, error);
+ return (error);
+ }
+ sc-sc_io_rman.rm_type = RMAN_ARRAY;
+ sc-sc_io_rman.rm_descr = CPCHT I/O Memory;
+ error = rman_init(sc-sc_io_rman);
if (error) {
device_printf(dev, rman_init() failed. error = %d\n, error);
return (error);
@@ -227,6 +238,9 @@ cpcht_attach(device_t dev)
* where we get the HT interrupts properties.
*/
+ /* I/O port mappings are usually not in the device tree */
+ rman_manage_region(sc-sc_io_rman, 0, CPCHT_IOPORT_SIZE - 1);
+
bzero(sc-htirq_map, sizeof(sc-htirq_map));
for (child = OF_child(node); child != 0; child = OF_peer(child))
cpcht_configure_htbridge(dev, child);
@@ -275,6 +289,9 @@ cpcht_configure_htbridge(device_t dev, p
case OFW_PCI_PHYS_HI_SPACE_CONFIG:
break;
case OFW_PCI_PHYS_HI_SPACE_IO:
+ rman_manage_region(sc-sc_io_rman, rp-pci_lo,
+ rp-pci_lo + rp-size_lo - 1);
+ break;
case OFW_PCI_PHYS_HI_SPACE_MEM32:
rman_manage_region(sc-sc_mem_rman, rp-pci_lo,
rp-pci_lo + rp-size_lo - 1);
@@ -482,8 +499,9 @@ cpcht_alloc_resource(device_t bus, devic
switch (type) {
case SYS_RES_IOPORT:
end = min(end, start + count);
+ rm = sc-sc_io_rman;
+ break;
- /* FALLTHROUGH */
case SYS_RES_MEMORY:
rm = sc-sc_mem_rman;
break;
@@ -537,6 +555,9 @@ cpcht_activate_resource(device_t bus, de
start = (vm_offset_t)rman_get_start(res);
+ if (type == SYS_RES_IOPORT)
+ start += CPCHT_IOPORT_BASE;
+
if (bootverbose)
printf(cpcht mapdev: start %zx, len %ld\n, start,
rman_get_size(res));
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215691 - stable/8/sys/powerpc/powerpc
Author: nwhitehorn
Date: Mon Nov 22 17:06:24 2010
New Revision: 215691
URL: http://svn.freebsd.org/changeset/base/215691
Log:
MFC r204127,215101:
Provide support for the DEEPNAP power-saving mode found on the 970MP.
Modified:
stable/8/sys/powerpc/powerpc/cpu.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/powerpc/cpu.c
==
--- stable/8/sys/powerpc/powerpc/cpu.c Mon Nov 22 17:03:09 2010
(r215690)
+++ stable/8/sys/powerpc/powerpc/cpu.c Mon Nov 22 17:06:24 2010
(r215691)
@@ -453,8 +453,16 @@ cpu_970_setup(int cpuid, uint16_t vers)
: =r (hid0_hi), =r (hid0_lo) : K (SPR_HID0));
/* Configure power-saving mode */
- hid0_hi |= (HID0_NAP | HID0_DPM);
- hid0_hi = ~(HID0_DOZE | HID0_DEEPNAP);
+ switch (vers) {
+ case IBM970MP:
+ hid0_hi |= (HID0_DEEPNAP | HID0_NAP | HID0_DPM);
+ hid0_hi = ~HID0_DOZE;
+ break;
+ default:
+ hid0_hi |= (HID0_NAP | HID0_DPM);
+ hid0_hi = ~(HID0_DOZE | HID0_DEEPNAP);
+ break;
+ }
powerpc_pow_enabled = 1;
__asm __volatile ( \
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215692 - stable/8/sys/powerpc/powermac
Author: nwhitehorn
Date: Mon Nov 22 17:09:42 2010
New Revision: 215692
URL: http://svn.freebsd.org/changeset/base/215692
Log:
MFC r215100:
Disabling CPU NAP modes during SMU commands is a hack needed only on U3
systems. Don't use it on non-U3 systems to allow cpu_idle() to work
correctly.
Modified:
stable/8/sys/powerpc/powermac/smu.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/powermac/smu.c
==
--- stable/8/sys/powerpc/powermac/smu.c Mon Nov 22 17:06:24 2010
(r215691)
+++ stable/8/sys/powerpc/powermac/smu.c Mon Nov 22 17:09:42 2010
(r215692)
@@ -89,6 +89,7 @@ struct smu_softc {
struct resource *sc_memr;
int sc_memrid;
+ int sc_u3;
bus_dma_tag_t sc_dmatag;
bus_space_tag_t sc_bt;
@@ -243,6 +244,10 @@ smu_attach(device_t dev)
sc-sc_cur_cmd = NULL;
sc-sc_doorbellirqid = -1;
+ sc-sc_u3 = 0;
+ if (OF_finddevice(/u3) != -1)
+ sc-sc_u3 = 1;
+
/*
* Map the mailbox area. This should be determined from firmware,
* but I have not found a simple way to do that.
@@ -361,7 +366,9 @@ smu_send_cmd(device_t dev, struct smu_cm
mtx_assert(sc-sc_mtx, MA_OWNED);
- powerpc_pow_enabled = 0;/* SMU cannot work if we go to NAP */
+ if (sc-sc_u3)
+ powerpc_pow_enabled = 0; /* SMU cannot work if we go to NAP */
+
sc-sc_cur_cmd = cmd;
/* Copy the command to the mailbox */
@@ -408,7 +415,8 @@ smu_doorbell_intr(void *xdev)
sizeof(sc-sc_cmd-data));
wakeup(sc-sc_cur_cmd);
sc-sc_cur_cmd = NULL;
- powerpc_pow_enabled = 1;
+ if (sc-sc_u3)
+ powerpc_pow_enabled = 1;
done:
/* Queue next command if one is pending */
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215693 - in stable/8/sys/powerpc: aim powermac
Author: nwhitehorn
Date: Mon Nov 22 17:13:04 2010
New Revision: 215693
URL: http://svn.freebsd.org/changeset/base/215693
Log:
MFC r205506:
Get nexus(4) out of the RTC business. The interface used by nexus(4)
in Open Firmware was Apple-specific, and we have complete coverage of Apple
system controllers, so move RTC responsibilities into the system controller
drivers. This avoids interesting problems from manipulating these devices
through Open Firmware behind the backs of their drivers.
Obtained from:NetBSD
Modified:
stable/8/sys/powerpc/aim/nexus.c
stable/8/sys/powerpc/powermac/cuda.c
stable/8/sys/powerpc/powermac/cudavar.h
stable/8/sys/powerpc/powermac/pmu.c
stable/8/sys/powerpc/powermac/smu.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/aim/nexus.c
==
--- stable/8/sys/powerpc/aim/nexus.cMon Nov 22 17:09:42 2010
(r215692)
+++ stable/8/sys/powerpc/aim/nexus.cMon Nov 22 17:13:04 2010
(r215693)
@@ -60,7 +60,6 @@
#include sys/systm.h
#include sys/module.h
#include sys/bus.h
-#include sys/clock.h
#include sys/cons.h
#include sys/kernel.h
#include sys/malloc.h
@@ -74,7 +73,6 @@
#include sys/rman.h
-#include clock_if.h
#include ofw_bus_if.h
#include pic_if.h
@@ -143,12 +141,6 @@ static const char *nexus_ofw_get_type(de
static const char *nexus_ofw_get_compat(device_t, device_t);
/*
- * Clock interface.
- */
-static int nexus_gettime(device_t, struct timespec *);
-static int nexus_settime(device_t, struct timespec *);
-
-/*
* Local routines
*/
static device_tnexus_device_from_node(device_t, phandle_t);
@@ -181,10 +173,6 @@ static device_method_t nexus_methods[] =
DEVMETHOD(ofw_bus_get_type, nexus_ofw_get_type),
DEVMETHOD(ofw_bus_get_compat, nexus_ofw_get_compat),
- /* Clock interface */
- DEVMETHOD(clock_gettime,nexus_gettime),
- DEVMETHOD(clock_settime,nexus_settime),
-
{ 0, 0 }
};
@@ -240,7 +228,6 @@ nexus_attach(device_t dev)
}
- clock_register(dev, 1000);
return (bus_generic_attach(dev));
}
@@ -512,50 +499,3 @@ nexus_ofw_get_compat(device_t bus, devic
return (dinfo-ndi_compatible);
}
-#defineDIFF190419702082844800
-
-static int
-nexus_gettime(device_t dev, struct timespec *ts)
-{
- char path[128];
- ihandle_t ih;
- phandle_t ph;
- u_int rtc;
-
- ph = OF_finddevice(rtc);
- if (ph == -1)
- return (ENOENT);
-
- OF_package_to_path(ph, path, sizeof(path));
- ih = OF_open(path);
- if (ih == -1)
- return (ENXIO);
-
- if (OF_call_method(read-rtc, ih, 0, 1, rtc))
- return (EIO);
-
- ts-tv_sec = rtc - DIFF19041970;
- ts-tv_nsec = 0;
- return (0);
-}
-
-static int
-nexus_settime(device_t dev, struct timespec *ts)
-{
- char path[128];
- ihandle_t ih;
- phandle_t ph;
- u_int rtc;
-
- ph = OF_finddevice(rtc);
- if (ph == -1)
- return (ENOENT);
-
- OF_package_to_path(ph, path, sizeof(path));
- ih = OF_open(path);
- if (ih == -1)
- return (ENXIO);
-
- rtc = ts-tv_sec + DIFF19041970;
- return ((OF_call_method(write-rtc, ih, 1, 0, rtc) != 0) ? EIO : 0);
-}
Modified: stable/8/sys/powerpc/powermac/cuda.c
==
--- stable/8/sys/powerpc/powermac/cuda.cMon Nov 22 17:09:42 2010
(r215692)
+++ stable/8/sys/powerpc/powermac/cuda.cMon Nov 22 17:13:04 2010
(r215693)
@@ -37,6 +37,7 @@ __FBSDID($FreeBSD$);
#include sys/bus.h
#include sys/conf.h
#include sys/kernel.h
+#include sys/clock.h
#include dev/ofw/ofw_bus.h
#include dev/ofw/openfirm.h
@@ -54,6 +55,7 @@ __FBSDID($FreeBSD$);
#include dev/adb/adb.h
+#include clock_if.h
#include cudavar.h
#include viareg.h
@@ -71,6 +73,12 @@ static u_int cuda_poll(device_t dev);
static voidcuda_send_inbound(struct cuda_softc *sc);
static voidcuda_send_outbound(struct cuda_softc *sc);
+/*
+ * Clock interface
+ */
+static int cuda_gettime(device_t dev, struct timespec *ts);
+static int cuda_settime(device_t dev, struct timespec *ts);
+
static device_method_t cuda_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, cuda_probe),
@@ -89,6 +97,10 @@ static device_method_t cuda_methods[] =
DEVMETHOD(adb_hb_controller_poll, cuda_poll),
DEVMETHOD(adb_hb_set_autopoll_mask, cuda_adb_autopoll),
+
svn commit: r215694 - stable/8/sys/powerpc/powermac
Author: nwhitehorn
Date: Mon Nov 22 17:14:29 2010
New Revision: 215694
URL: http://svn.freebsd.org/changeset/base/215694
Log:
MFC r208841:
Add support for the I2C busses hanging off Apple system management chips.
Modified:
stable/8/sys/powerpc/powermac/smu.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/powerpc/powermac/smu.c
==
--- stable/8/sys/powerpc/powermac/smu.c Mon Nov 22 17:13:04 2010
(r215693)
+++ stable/8/sys/powerpc/powermac/smu.c Mon Nov 22 17:14:29 2010
(r215694)
@@ -47,12 +47,16 @@ __FBSDID($FreeBSD$);
#include machine/intr_machdep.h
#include machine/md_var.h
+#include dev/iicbus/iicbus.h
+#include dev/iicbus/iiconf.h
#include dev/led/led.h
#include dev/ofw/openfirm.h
#include dev/ofw/ofw_bus.h
+#include dev/ofw/ofw_bus_subr.h
#include powerpc/powermac/macgpiovar.h
#include clock_if.h
+#include iicbus_if.h
struct smu_cmd {
volatile uint8_t cmd;
@@ -138,6 +142,8 @@ struct smu_softc {
static int smu_probe(device_t);
static int smu_attach(device_t);
+static const struct ofw_bus_devinfo *
+smu_get_devinfo(device_t bus, device_t dev);
/* cpufreq notification hooks */
@@ -152,6 +158,7 @@ static int smu_settime(device_t dev, str
static int smu_run_cmd(device_t dev, struct smu_cmd *cmd, int wait);
static int smu_get_datablock(device_t dev, int8_t id, uint8_t *buf,
size_t len);
+static voidsmu_attach_i2c(device_t dev, phandle_t i2croot);
static voidsmu_attach_fans(device_t dev, phandle_t fanroot);
static voidsmu_attach_sensors(device_t dev, phandle_t sensroot);
static voidsmu_fan_management_proc(void *xdev);
@@ -172,6 +179,16 @@ static device_method_t smu_methods[] =
/* Clock interface */
DEVMETHOD(clock_gettime,smu_gettime),
DEVMETHOD(clock_settime,smu_settime),
+
+ /* ofw_bus interface */
+ DEVMETHOD(bus_child_pnpinfo_str,ofw_bus_gen_child_pnpinfo_str),
+ DEVMETHOD(ofw_bus_get_devinfo, smu_get_devinfo),
+ DEVMETHOD(ofw_bus_get_compat, ofw_bus_gen_get_compat),
+ DEVMETHOD(ofw_bus_get_model,ofw_bus_gen_get_model),
+ DEVMETHOD(ofw_bus_get_name, ofw_bus_gen_get_name),
+ DEVMETHOD(ofw_bus_get_node, ofw_bus_gen_get_node),
+ DEVMETHOD(ofw_bus_get_type, ofw_bus_gen_get_type),
+
{ 0, 0 },
};
@@ -305,8 +322,14 @@ smu_attach(device_t dev)
if (strncmp(name, sensors, 8) == 0)
smu_attach_sensors(dev, child);
+
+ if (strncmp(name, smu-i2c-control, 15) == 0)
+ smu_attach_i2c(dev, child);
}
+ /* Some SMUs have the I2C children directly under the bus. */
+ smu_attach_i2c(dev, node);
+
/*
* Collect calibration constants.
*/
@@ -373,7 +396,14 @@ smu_attach(device_t dev)
*/
clock_register(dev, 1000);
- return (0);
+ return (bus_generic_attach(dev));
+}
+
+static const struct ofw_bus_devinfo *
+smu_get_devinfo(device_t bus, device_t dev)
+{
+
+ return (device_get_ivars(dev));
}
static void
@@ -795,8 +825,8 @@ smu_attach_fans(device_t dev, phandle_t
CTLTYPE_INT | CTLFLAG_RD, fan-max_rpm, sizeof(cell_t),
Maximum allowed RPM);
SYSCTL_ADD_PROC(ctx, SYSCTL_CHILDREN(oid), OID_AUTO, rpm,
- CTLTYPE_INT | CTLFLAG_RW, dev, sc-sc_nfans,
- smu_fanrpm_sysctl, I, Fan RPM);
+ CTLTYPE_INT | CTLFLAG_RW | CTLFLAG_MPSAFE, dev,
+ sc-sc_nfans, smu_fanrpm_sysctl, I, Fan RPM);
fan++;
sc-sc_nfans++;
@@ -959,8 +989,8 @@ smu_attach_sensors(device_t dev, phandle
sprintf(sysctl_desc,%s (%s), sens-location, units);
SYSCTL_ADD_PROC(ctx, SYSCTL_CHILDREN(sensroot_oid), OID_AUTO,
- sysctl_name, CTLTYPE_INT | CTLFLAG_RD, dev, sc-sc_nsensors,
- smu_sensor_sysctl, I, sysctl_desc);
+ sysctl_name, CTLTYPE_INT | CTLFLAG_RD | CTLFLAG_MPSAFE,
+ dev, sc-sc_nsensors, smu_sensor_sysctl, I, sysctl_desc);
sens++;
sc-sc_nsensors++;
@@ -996,13 +1026,6 @@ smu_manage_fans(device_t smu)
maxtemp = temp;
}
- if (maxtemp 10) { /* Bail if no good sensors */
- for (i = 0; i sc-sc_nfans; i++)
- smu_fan_set_rpm(smu, sc-sc_fans[i],
- sc-sc_fans[i].unmanaged_rpm);
- return;
- }
-
if (maxtemp
svn commit: r215695 - in stable/8/sys: conf powerpc/powermac
Author: nwhitehorn
Date: Mon Nov 22 17:15:41 2010
New Revision: 215695
URL: http://svn.freebsd.org/changeset/base/215695
Log:
MFC r208842:
Add a driver for the CPU temperature sensors attached over I2C on the
PowerMac 11,2.
Added:
stable/8/sys/powerpc/powermac/smusat.c
- copied unchanged from r208842, head/sys/powerpc/powermac/smusat.c
Modified:
stable/8/sys/conf/files.powerpc
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/conf/files.powerpc
==
--- stable/8/sys/conf/files.powerpc Mon Nov 22 17:14:29 2010
(r215694)
+++ stable/8/sys/conf/files.powerpc Mon Nov 22 17:15:41 2010
(r215695)
@@ -140,6 +140,7 @@ powerpc/powermac/openpic_macio.c optiona
powerpc/powermac/pswitch.c optionalpowermac pswitch
powerpc/powermac/pmu.c optionalpowermac pmu
powerpc/powermac/smu.c optionalpowermac smu
+powerpc/powermac/smusat.c optionalpowermac smu
powerpc/powermac/uninorth.coptionalpowermac
powerpc/powermac/uninorthpci.c optionalpowermac pci
powerpc/powermac/vcoregpio.c optionalpowermac
Copied: stable/8/sys/powerpc/powermac/smusat.c (from r208842,
head/sys/powerpc/powermac/smusat.c)
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ stable/8/sys/powerpc/powermac/smusat.c Mon Nov 22 17:15:41 2010
(r215695, copy of r208842, head/sys/powerpc/powermac/smusat.c)
@@ -0,0 +1,262 @@
+/*-
+ * Copyright (c) 2010 Nathan Whitehorn
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *notice, this list of conditions and the following disclaimer in the
+ *documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+ * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
+ * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
+ * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ *
+ */
+
+#include sys/cdefs.h
+__FBSDID($FreeBSD$);
+
+#include sys/param.h
+#include sys/systm.h
+#include sys/module.h
+#include sys/bus.h
+#include sys/conf.h
+#include sys/cpu.h
+#include sys/ctype.h
+#include sys/kernel.h
+#include sys/sysctl.h
+
+#include dev/iicbus/iicbus.h
+#include dev/iicbus/iiconf.h
+#include dev/ofw/ofw_bus.h
+#include dev/ofw/openfirm.h
+
+struct smu_sensor {
+ cell_t reg;
+ charlocation[32];
+ enum {
+ SMU_CURRENT_SENSOR,
+ SMU_VOLTAGE_SENSOR,
+ SMU_POWER_SENSOR,
+ SMU_TEMP_SENSOR
+ } type;
+};
+
+static int smusat_probe(device_t);
+static int smusat_attach(device_t);
+static int smusat_sensor_sysctl(SYSCTL_HANDLER_ARGS);
+
+MALLOC_DEFINE(M_SMUSAT, smusat, SMU Sattelite Sensors);
+
+static device_method_t smusat_methods[] = {
+ /* Device interface */
+ DEVMETHOD(device_probe, smusat_probe),
+ DEVMETHOD(device_attach,smusat_attach),
+ { 0, 0 },
+};
+
+struct smusat_softc {
+ struct smu_sensor *sc_sensors;
+ int sc_nsensors;
+
+ uint8_t sc_cache[16];
+ time_t sc_last_update;
+};
+
+static driver_t smusat_driver = {
+ smusat,
+ smusat_methods,
+ sizeof(struct smusat_softc)
+};
+
+static devclass_t smusat_devclass;
+
+DRIVER_MODULE(smusat, iicbus, smusat_driver, smusat_devclass, 0, 0);
+
+static int
+smusat_probe(device_t dev)
+{
+ const char *compat = ofw_bus_get_compat(dev);
+
+ if (compat == NULL || strcmp(compat, smu-sat) != 0)
+ return (ENXIO);
+
+ device_set_desc(dev, SMU Satellite Sensors);
+ return (0);
+}
+
+static int
+smusat_attach(device_t dev)
+{
+ phandle_t
svn commit: r215696 - in stable/8/sys: conf powerpc/aim powerpc/booke powerpc/include powerpc/mpc85xx powerpc/powermac powerpc/powerpc
Author: nwhitehorn
Date: Mon Nov 22 17:39:18 2010
New Revision: 215696
URL: http://svn.freebsd.org/changeset/base/215696
Log:
MFC r212054:
Restructure how reset and poweroff are handled on PowerPC systems, since
the existing code was very platform specific, and broken for SMP systems
trying to reboot from KDB.
- Add a new PLATFORM_RESET() method to the platform KOBJ interface, and
migrate existing reset functions into platform modules.
- Modify the OF_reboot() routine to submit the request by hand to avoid
the IPIs involved in the regular openfirmware() routine. This fixes
reboot from KDB on SMP machines.
- Move non-KDB reset and poweroff functions on the Powermac platform
into the relevant power control drivers (cuda, pmu, smu), instead of
using them through the Open Firmware backdoor.
- Rename platform_chrp to platform_powermac since it has become
increasingly Powermac specific. When we gain support for IBM systems,
we will grow a new platform_chrp.
Added:
stable/8/sys/powerpc/powermac/platform_powermac.c
- copied unchanged from r212054,
head/sys/powerpc/powermac/platform_powermac.c
Modified:
stable/8/sys/conf/files.powerpc
stable/8/sys/powerpc/aim/machdep.c
stable/8/sys/powerpc/aim/ofw_machdep.c
stable/8/sys/powerpc/aim/vm_machdep.c
stable/8/sys/powerpc/booke/platform_bare.c
stable/8/sys/powerpc/include/ofw_machdep.h
stable/8/sys/powerpc/mpc85xx/mpc85xx.c
stable/8/sys/powerpc/powermac/cuda.c
stable/8/sys/powerpc/powermac/pmu.c
stable/8/sys/powerpc/powermac/smu.c
stable/8/sys/powerpc/powerpc/platform.c
stable/8/sys/powerpc/powerpc/platform_if.m
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/conf/files.powerpc
==
--- stable/8/sys/conf/files.powerpc Mon Nov 22 17:15:41 2010
(r215695)
+++ stable/8/sys/conf/files.powerpc Mon Nov 22 17:39:18 2010
(r215696)
@@ -82,7 +82,6 @@ powerpc/aim/mp_cpudep.c optionalaim sm
powerpc/aim/nexus.coptionalaim
powerpc/aim/ofw_machdep.c optionalaim
powerpc/aim/ofwmagic.S optionalaim
-powerpc/aim/platform_chrp.coptionalaim
powerpc/aim/swtch.Soptionalaim
powerpc/aim/trap.c optionalaim
powerpc/aim/uma_machdep.c optionalaim
@@ -137,6 +136,7 @@ powerpc/powermac/kiic.c optionalpowerm
powerpc/powermac/macgpio.c optionalpowermac pci
powerpc/powermac/macio.c optionalpowermac pci
powerpc/powermac/openpic_macio.c optional powermac pci
+powerpc/powermac/platform_powermac.c optional powermac
powerpc/powermac/pswitch.c optionalpowermac pswitch
powerpc/powermac/pmu.c optionalpowermac pmu
powerpc/powermac/smu.c optionalpowermac smu
Modified: stable/8/sys/powerpc/aim/machdep.c
==
--- stable/8/sys/powerpc/aim/machdep.c Mon Nov 22 17:15:41 2010
(r215695)
+++ stable/8/sys/powerpc/aim/machdep.c Mon Nov 22 17:39:18 2010
(r215696)
@@ -167,15 +167,6 @@ struct bat battable[16];
struct kva_md_info kmi;
static void
-powerpc_ofw_shutdown(void *junk, int howto)
-{
- if (howto RB_HALT) {
- OF_halt();
- }
- OF_reboot();
-}
-
-static void
cpu_startup(void *dummy)
{
@@ -227,9 +218,6 @@ cpu_startup(void *dummy)
*/
bufinit();
vm_pager_bufferinit();
-
- EVENTHANDLER_REGISTER(shutdown_final, powerpc_ofw_shutdown, 0,
- SHUTDOWN_PRI_LAST);
}
extern charkernel_text[], _end[];
Modified: stable/8/sys/powerpc/aim/ofw_machdep.c
==
--- stable/8/sys/powerpc/aim/ofw_machdep.c Mon Nov 22 17:15:41 2010
(r215695)
+++ stable/8/sys/powerpc/aim/ofw_machdep.c Mon Nov 22 17:39:18 2010
(r215696)
@@ -463,20 +463,21 @@ openfirmware(void *args)
}
void
-OF_halt()
-{
- int retval; /* dummy, this may not be needed */
-
- OF_interpret(shut-down, 1, retval);
- for (;;); /* just in case */
-}
-
-void
OF_reboot()
{
- int retval; /* dummy, this may not be needed */
+ struct {
+ cell_t name;
+ cell_t nargs;
+ cell_t nreturns;
+ cell_t arg;
+ } args;
+
+ args.name = (cell_t)(uintptr_t)interpret;
+ args.nargs = 1;
+ args.nreturns = 0;
+ args.arg = (cell_t)(uintptr_t)reset-all;
+ openfirmware_core(args); /* Don't do
Re: svn commit: r215679 - in head: contrib/binutils/bfd contrib/binutils/binutils contrib/binutils/include/elf contrib/file contrib/gdb/gdb gnu/usr.bin/gdb/libgdb sys/compat/freebsd32 sys/kern sys/sys
On Mon, Nov 22, 2010 at 02:42:13PM +, Attilio Rao wrote: Author: attilio Date: Mon Nov 22 14:42:13 2010 New Revision: 215679 URL: http://svn.freebsd.org/changeset/base/215679 Log: Add the ability for GDB to printout the thread name along with other thread specific informations. This change is particularly useful if one has a heavily threaded app that uses pthread_set_name_np() to set its thread names. This is an example of the output you get with stock GDB: (gdb) info threads 12 Thread 2a4041c0 (LWP 100184) 0x29e91fe3 in memset () 11 Thread 2a404700 (LWP 100190) 0x29e2be33 in poll () 10 Thread 2a404c40 (LWP 100191) 0x29ed85e7 in _umtx_op_err () And with this change: (gdb) info threads 41 Thread 0x9d8 (LWP 100259/initial thread) 0x6a3f23ff in nanosleep () 40 Thread 0x9f48600 (LWP 100315/EventLoop) 0x6a38e427 in _umtx_op () 39 Thread 0x9f48800 (LWP 100321/ntfy) 0x6a38e427 in _umtx_op () -Ed ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215697 - in head/crypto/openssl: . apps crypto crypto/aes crypto/aes/asm crypto/asn1 crypto/bio crypto/bn crypto/cms crypto/conf crypto/des crypto/dsa crypto/ec crypto/ecdh crypto/ecds...
Author: simon Date: Mon Nov 22 18:23:44 2010 New Revision: 215697 URL: http://svn.freebsd.org/changeset/base/215697 Log: Merge OpenSSL 0.9.8p into head. Security: CVE-2010-3864 Security: http://www.openssl.org/news/secadv_20101116.txt Modified: head/crypto/openssl/CHANGES head/crypto/openssl/Configure head/crypto/openssl/FAQ head/crypto/openssl/Makefile head/crypto/openssl/NEWS head/crypto/openssl/PROBLEMS head/crypto/openssl/README head/crypto/openssl/apps/apps.c head/crypto/openssl/apps/dh.c head/crypto/openssl/apps/dhparam.c head/crypto/openssl/apps/dsaparam.c head/crypto/openssl/apps/ec.c head/crypto/openssl/apps/ecparam.c head/crypto/openssl/apps/enc.c head/crypto/openssl/apps/gendh.c head/crypto/openssl/apps/gendsa.c head/crypto/openssl/apps/genrsa.c head/crypto/openssl/apps/pkcs7.c head/crypto/openssl/apps/rand.c head/crypto/openssl/apps/s_server.c head/crypto/openssl/apps/s_socket.c head/crypto/openssl/apps/speed.c head/crypto/openssl/apps/x509.c head/crypto/openssl/crypto/aes/aes_wrap.c head/crypto/openssl/crypto/aes/asm/aes-x86_64.pl head/crypto/openssl/crypto/asn1/a_int.c head/crypto/openssl/crypto/asn1/n_pkey.c head/crypto/openssl/crypto/asn1/t_crl.c head/crypto/openssl/crypto/asn1/tasn_dec.c head/crypto/openssl/crypto/asn1/x_x509.c head/crypto/openssl/crypto/bio/b_sock.c head/crypto/openssl/crypto/bio/bf_nbio.c head/crypto/openssl/crypto/bio/bio_lib.c head/crypto/openssl/crypto/bio/bss_acpt.c head/crypto/openssl/crypto/bio/bss_sock.c head/crypto/openssl/crypto/bn/bn_exp2.c head/crypto/openssl/crypto/bn/bn_mul.c head/crypto/openssl/crypto/cms/cms_asn1.c head/crypto/openssl/crypto/conf/conf_def.c head/crypto/openssl/crypto/des/rpc_des.h head/crypto/openssl/crypto/dsa/dsa_gen.c head/crypto/openssl/crypto/dsa/dsa_ossl.c head/crypto/openssl/crypto/ec/ec2_mult.c head/crypto/openssl/crypto/ec/ec_mult.c head/crypto/openssl/crypto/ecdh/ech_lib.c head/crypto/openssl/crypto/ecdsa/ecs_lib.c head/crypto/openssl/crypto/engine/eng_list.c head/crypto/openssl/crypto/err/err_prn.c head/crypto/openssl/crypto/evp/bio_b64.c head/crypto/openssl/crypto/evp/enc_min.c head/crypto/openssl/crypto/evp/encode.c head/crypto/openssl/crypto/evp/evp_pbe.c head/crypto/openssl/crypto/hmac/hmac.c head/crypto/openssl/crypto/md32_common.h head/crypto/openssl/crypto/o_init.c head/crypto/openssl/crypto/ocsp/ocsp_ht.c head/crypto/openssl/crypto/ocsp/ocsp_prn.c head/crypto/openssl/crypto/opensslv.h head/crypto/openssl/crypto/pem/pem_lib.c head/crypto/openssl/crypto/pkcs12/p12_key.c head/crypto/openssl/crypto/pkcs12/p12_npas.c head/crypto/openssl/crypto/pkcs7/pk7_doit.c head/crypto/openssl/crypto/pkcs7/pk7_lib.c head/crypto/openssl/crypto/pkcs7/pk7_mime.c head/crypto/openssl/crypto/pqueue/pqueue.c head/crypto/openssl/crypto/rand/rand_nw.c head/crypto/openssl/crypto/rand/randfile.c head/crypto/openssl/crypto/rsa/rsa_eay.c head/crypto/openssl/crypto/x509/x509.h head/crypto/openssl/crypto/x509/x509_vfy.c head/crypto/openssl/crypto/x509/x_all.c head/crypto/openssl/crypto/x509v3/v3_ncons.c head/crypto/openssl/crypto/x509v3/v3_pci.c head/crypto/openssl/doc/apps/smime.pod head/crypto/openssl/doc/crypto/ASN1_OBJECT_new.pod head/crypto/openssl/doc/crypto/ASN1_STRING_length.pod head/crypto/openssl/doc/crypto/ASN1_STRING_new.pod head/crypto/openssl/doc/crypto/ASN1_generate_nconf.pod head/crypto/openssl/doc/crypto/BIO_f_buffer.pod head/crypto/openssl/doc/crypto/BIO_should_retry.pod head/crypto/openssl/doc/crypto/CRYPTO_set_ex_data.pod head/crypto/openssl/doc/crypto/OBJ_nid2obj.pod head/crypto/openssl/doc/crypto/PKCS7_decrypt.pod head/crypto/openssl/doc/crypto/PKCS7_encrypt.pod head/crypto/openssl/doc/crypto/PKCS7_sign.pod head/crypto/openssl/doc/crypto/PKCS7_verify.pod head/crypto/openssl/doc/crypto/SMIME_read_PKCS7.pod head/crypto/openssl/doc/crypto/SMIME_write_PKCS7.pod head/crypto/openssl/doc/crypto/X509_NAME_ENTRY_get_object.pod head/crypto/openssl/doc/crypto/X509_NAME_add_entry_by_txt.pod head/crypto/openssl/doc/crypto/X509_NAME_get_index_by_NID.pod head/crypto/openssl/doc/crypto/X509_new.pod head/crypto/openssl/doc/crypto/bn_internal.pod head/crypto/openssl/doc/crypto/ui_compat.pod head/crypto/openssl/doc/ssl/SSL_library_init.pod head/crypto/openssl/e_os.h head/crypto/openssl/engines/e_chil.c head/crypto/openssl/engines/e_cswift.c head/crypto/openssl/engines/e_ubsec.c head/crypto/openssl/fips/mkfipsscr.pl head/crypto/openssl/openssl.spec head/crypto/openssl/ssl/d1_both.c head/crypto/openssl/ssl/d1_clnt.c head/crypto/openssl/ssl/d1_enc.c head/crypto/openssl/ssl/d1_lib.c head/crypto/openssl/ssl/d1_pkt.c head/crypto/openssl/ssl/dtls1.h head/crypto/openssl/ssl/s23_clnt.c head/crypto/openssl/ssl/s23_lib.c head/crypto/openssl/ssl/s2_srvr.c head/crypto/openssl/ssl/s3_both.c head/crypto/openssl/ssl/s3_clnt.c
svn commit: r215699 - head/sys/net80211
Author: bschmidt Date: Mon Nov 22 19:01:47 2010 New Revision: 215699 URL: http://svn.freebsd.org/changeset/base/215699 Log: The meshid element is memcpy()'ed into se_meshid if included in either beacon or probe-response frames. Fix the condition by checking for the the array's content instead of the always existing array itself. Reviewed by: rpaulo, stefanf MFC after:3 days Modified: head/sys/net80211/ieee80211_scan_sta.c Modified: head/sys/net80211/ieee80211_scan_sta.c == --- head/sys/net80211/ieee80211_scan_sta.c Mon Nov 22 18:29:00 2010 (r215698) +++ head/sys/net80211/ieee80211_scan_sta.c Mon Nov 22 19:01:47 2010 (r215699) @@ -1013,7 +1013,7 @@ match_bss(struct ieee80211vap *vap, */ if (se-se_capinfo (IEEE80211_CAPINFO_IBSS|IEEE80211_CAPINFO_ESS)) fail |= MATCH_CAPINFO; - else if (se-se_meshid == NULL) + else if (se-se_meshid[0] != IEEE80211_ELEMID_MESHID) fail |= MATCH_MESH_NOID; else if (ms-ms_idlen != 0 match_id(se-se_meshid, ms-ms_id, ms-ms_idlen)) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215700 - in stable/8: sbin/iscontrol sys/dev/iscsi/initiator
Author: des
Date: Mon Nov 22 19:02:30 2010
New Revision: 215700
URL: http://svn.freebsd.org/changeset/base/215700
Log:
MFC r209052, r210702, r211095, r211182, r212149: Fix a memory leak and
some potential deadlocks, increase the target limit from 4 to 64, and
numerous other scalability and stability improvements, including.
Submitted by: Daniel Braniss da...@cs.huji.ac.il
Sponsored by: Dansk Scanning A/S, Data Robotics Inc.
Deleted:
stable/8/sbin/iscontrol/pdu.h
Modified:
stable/8/sbin/iscontrol/Makefile
stable/8/sbin/iscontrol/auth_subr.c
stable/8/sbin/iscontrol/config.c
stable/8/sbin/iscontrol/fsm.c
stable/8/sbin/iscontrol/iscontrol.8
stable/8/sbin/iscontrol/iscontrol.c
stable/8/sbin/iscontrol/iscontrol.h
stable/8/sbin/iscontrol/iscsi.conf.5
stable/8/sbin/iscontrol/login.c
stable/8/sbin/iscontrol/misc.c
stable/8/sbin/iscontrol/pdu.c
stable/8/sys/dev/iscsi/initiator/isc_cam.c
stable/8/sys/dev/iscsi/initiator/isc_sm.c
stable/8/sys/dev/iscsi/initiator/isc_soc.c
stable/8/sys/dev/iscsi/initiator/isc_subr.c
stable/8/sys/dev/iscsi/initiator/iscsi.c
stable/8/sys/dev/iscsi/initiator/iscsi.h
stable/8/sys/dev/iscsi/initiator/iscsi_subr.c
stable/8/sys/dev/iscsi/initiator/iscsivar.h
Directory Properties:
stable/8/sbin/iscontrol/ (props changed)
Modified: stable/8/sbin/iscontrol/Makefile
==
--- stable/8/sbin/iscontrol/MakefileMon Nov 22 19:01:47 2010
(r215699)
+++ stable/8/sbin/iscontrol/MakefileMon Nov 22 19:02:30 2010
(r215700)
@@ -4,8 +4,10 @@ SRCS= iscontrol.c pdu.c fsm.c config.c l
PROG= iscontrol
DPADD= ${LIBCAM} ${LIBMD}
LDADD= -lcam -lmd
+S= ${.CURDIR}/../../sys
-CFLAGS += -I${.CURDIR}/../../sys/dev/iscsi/initiator
+WARNS?=3
+CFLAGS += -I$S
#CFLAGS += -g -DDEBUG
MAN= iscsi.conf.5 iscontrol.8
Modified: stable/8/sbin/iscontrol/auth_subr.c
==
--- stable/8/sbin/iscontrol/auth_subr.c Mon Nov 22 19:01:47 2010
(r215699)
+++ stable/8/sbin/iscontrol/auth_subr.c Mon Nov 22 19:02:30 2010
(r215700)
@@ -1,5 +1,5 @@
/*-
- * Copyright (c) 2005-2008 Daniel Braniss da...@cs.huji.ac.il
+ * Copyright (c) 2005-2010 Daniel Braniss da...@cs.huji.ac.il
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -52,7 +52,7 @@ __FBSDID($FreeBSD$);
#include md5.h
#include sha.h
-#include iscsi.h
+#include dev/iscsi/initiator/iscsi.h
#include iscontrol.h
static int
@@ -152,7 +152,7 @@ chapDigest(char *ap, char id, char *cp,
}
char *
-genChapChallenge(char *encoding, int len)
+genChapChallenge(char *encoding, uint len)
{
int fd;
unsigned char tmp[1024];
Modified: stable/8/sbin/iscontrol/config.c
==
--- stable/8/sbin/iscontrol/config.cMon Nov 22 19:01:47 2010
(r215699)
+++ stable/8/sbin/iscontrol/config.cMon Nov 22 19:02:30 2010
(r215700)
@@ -1,5 +1,5 @@
/*-
- * Copyright (c) 2005-2008 Daniel Braniss da...@cs.huji.ac.il
+ * Copyright (c) 2005-2009 Daniel Braniss da...@cs.huji.ac.il
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -41,7 +41,7 @@ __FBSDID($FreeBSD$);
#include ctype.h
#include camlib.h
-#include iscsi.h
+#include dev/iscsi/initiator/iscsi.h
#include iscontrol.h
/*
@@ -94,6 +94,11 @@ __FBSDID($FreeBSD$);
#define OPT_iqn34
#define OPT_sockbufsize35
+/*
+ | sentinel
+ */
+#define OPT_end0
+
#define _OFF(v)((int)((isc_opt_t *)NULL)-v)
#define _E(u, s, v) {.usage=u, .scope=s, .name=#v, .tokenID=OPT_##v}
@@ -145,7 +150,7 @@ textkey_t keyMap[] = {
_E(U_LO, S_SW, sessionType),
- {0}
+ _E(0, 0, end)
};
#define _OPT_INT(w)strtol((char *)w, NULL, 0)
@@ -154,7 +159,7 @@ textkey_t keyMap[] = {
static __inline int
_OPT_BOOL(char *w)
{
- if(isalpha(*w))
+ if(isalpha((unsigned char)*w))
return strcasecmp(w, TRUE) == 0;
else
return _OPT_INT(w);
@@ -244,13 +249,12 @@ getConfig(FILE *fd, char *key, char **Ar
len = 0;
state = 0;
while((lp = getline(fd)) != NULL) {
- for(; isspace(*lp); lp++)
+ for(; isspace((unsigned char)*lp); lp++)
;
switch(state) {
case 0:
if((p = strchr(lp, '{')) != NULL) {
- n = 0;
- while((--p lp) *p isspace(*p));
+ while((--p lp) *p isspace((unsigned char)*p));
n = p - lp;
if(len strncmp(lp, key, MAX(n, len)) == 0)
state = 2;
@@ -273,7 +277,7 @@ getConfig(FILE *fd, char *key, char **Ar
}
- for(p =
Re: svn commit: r215675 - head/sys/compat/linux
Den 22/11/2010 kl. 13.42 skrev Alexander Leidinger:
Author: netchild
Date: Mon Nov 22 12:42:32 2010
New Revision: 215675
URL: http://svn.freebsd.org/changeset/base/215675
Log:
Do not take the process lock. The assignment to u_short inside the
properly aligned structure is atomic on all supported architectures, and
the thread that should see side-effect of assignment is the same thread
that does assignment.
Use a more appropriate conditional to detect the linux ABI.
Suggested by:kib
X-MFC: together with r215664
Modified:
head/sys/compat/linux/linux_emul.c
Modified: head/sys/compat/linux/linux_emul.c
==
--- head/sys/compat/linux/linux_emul.cMon Nov 22 12:33:48 2010
(r215674)
+++ head/sys/compat/linux/linux_emul.cMon Nov 22 12:42:32 2010
(r215675)
@@ -198,11 +198,8 @@ linux_proc_exit(void *arg __unused, stru
} else
EMUL_SHARED_WUNLOCK(emul_shared_lock);
- if ((shared_flags EMUL_SHARED_HASXSTAT) != 0) {
- PROC_LOCK(p);
+ if ((shared_flags EMUL_SHARED_HASXSTAT) != 0)
p-p_xstat = shared_xstat;
- PROC_UNLOCK(p);
- }
if (child_clear_tid != NULL) {
struct linux_sys_futex_args cup;
@@ -265,7 +262,8 @@ linux_proc_exec(void *arg __unused, stru
if (__predict_false(imgp-sysent == elf_linux_sysvec
p-p_sysent != elf_linux_sysvec))
linux_proc_init(FIRST_THREAD_IN_PROC(p), p-p_pid, 0);
- if (__predict_false(p-p_sysent == elf_linux_sysvec))
+ if (__predict_false((p-p_sysent-sv_flags SV_ABI_MASK) ==
+ SV_ABI_LINUX))
/* Kill threads regardless of imgp-sysent value */
linux_kill_threads(FIRST_THREAD_IN_PROC(p), SIGKILL);
if (__predict_false(imgp-sysent != elf_linux_sysvec
This breaks make buildkernel for me (8.1 RELEASE host):
/usr/home/erik/freebsd/head/src/sys/modules/linux/../../compat/linux/linux_emul.c:
In function 'linux_proc_exec':
/usr/home/erik/freebsd/head/src/sys/modules/linux/../../compat/linux/linux_emul.c:265:
error: dereferencing pointer to incomplete type
/usr/home/erik/freebsd/head/src/sys/modules/linux/../../compat/linux/linux_emul.c:265:
error: 'SV_ABI_MASK' undeclared (first use in this function)
Erik
svn commit: r215701 - in head/sys: kern mips/mips mips/rmi net netgraph netinet netinet/ipfw netinet6 netipsec powerpc/aim powerpc/booke sys
Author: dim
Date: Mon Nov 22 19:32:54 2010
New Revision: 215701
URL: http://svn.freebsd.org/changeset/base/215701
Log:
After some off-list discussion, revert a number of changes to the
DPCPU_DEFINE and VNET_DEFINE macros, as these cause problems for various
people working on the affected files. A better long-term solution is
still being considered. This reversal may give some modules empty
set_pcpu or set_vnet sections, but these are harmless.
Changes reverted:
r215318 | dim | 2010-11-14 21:40:55 +0100 (Sun, 14 Nov 2010) | 4 lines
Instead of unconditionally emitting .globl's for the __start_set_xxx and
__stop_set_xxx symbols, only emit them when the set_vnet or set_pcpu
sections are actually defined.
r215317 | dim | 2010-11-14 21:38:11 +0100 (Sun, 14 Nov 2010) | 3 lines
Apply the STATIC_VNET_DEFINE and STATIC_DPCPU_DEFINE macros throughout
the tree.
r215316 | dim | 2010-11-14 21:23:02 +0100 (Sun, 14 Nov 2010) | 2 lines
Add macros to define static instances of VNET_DEFINE and DPCPU_DEFINE.
Modified:
head/sys/kern/kern_clock.c
head/sys/kern/kern_clocksource.c
head/sys/kern/sched_4bsd.c
head/sys/kern/subr_pcpu.c
head/sys/mips/mips/tick.c
head/sys/mips/rmi/tick.c
head/sys/net/flowtable.c
head/sys/net/if.c
head/sys/net/if_clone.c
head/sys/net/if_ethersubr.c
head/sys/net/if_gif.c
head/sys/net/if_llatbl.c
head/sys/net/if_loop.c
head/sys/net/route.c
head/sys/net/vnet.c
head/sys/net/vnet.h
head/sys/netgraph/ng_base.c
head/sys/netgraph/ng_eiface.c
head/sys/netgraph/ng_iface.c
head/sys/netinet/if_ether.c
head/sys/netinet/igmp.c
head/sys/netinet/in.c
head/sys/netinet/in_pcb.c
head/sys/netinet/in_rmx.c
head/sys/netinet/ip_divert.c
head/sys/netinet/ip_fastfwd.c
head/sys/netinet/ip_icmp.c
head/sys/netinet/ip_input.c
head/sys/netinet/ip_ipsec.c
head/sys/netinet/ip_mroute.c
head/sys/netinet/ipfw/ip_fw2.c
head/sys/netinet/ipfw/ip_fw_dynamic.c
head/sys/netinet/ipfw/ip_fw_nat.c
head/sys/netinet/ipfw/ip_fw_pfil.c
head/sys/netinet/siftr.c
head/sys/netinet/tcp_hostcache.c
head/sys/netinet/tcp_reass.c
head/sys/netinet/tcp_subr.c
head/sys/netinet/tcp_syncache.c
head/sys/netinet/tcp_timewait.c
head/sys/netinet/udp_usrreq.c
head/sys/netinet6/frag6.c
head/sys/netinet6/icmp6.c
head/sys/netinet6/in6_rmx.c
head/sys/netinet6/in6_src.c
head/sys/netinet6/ip6_ipsec.c
head/sys/netinet6/ip6_mroute.c
head/sys/netinet6/mld6.c
head/sys/netinet6/nd6.c
head/sys/netinet6/nd6_nbr.c
head/sys/netinet6/nd6_rtr.c
head/sys/netinet6/scope6.c
head/sys/netinet6/send.c
head/sys/netipsec/key.c
head/sys/netipsec/keysock.c
head/sys/netipsec/xform_esp.c
head/sys/powerpc/aim/clock.c
head/sys/powerpc/booke/clock.c
head/sys/sys/cdefs.h
head/sys/sys/linker_set.h
head/sys/sys/pcpu.h
Modified: head/sys/kern/kern_clock.c
==
--- head/sys/kern/kern_clock.c Mon Nov 22 19:02:30 2010(r215700)
+++ head/sys/kern/kern_clock.c Mon Nov 22 19:32:54 2010(r215701)
@@ -373,7 +373,7 @@ int profprocs;
intticks;
intpsratio;
-STATIC_DPCPU_DEFINE(int, pcputicks); /* Per-CPU version of ticks. */
+static DPCPU_DEFINE(int, pcputicks); /* Per-CPU version of ticks. */
static int global_hardclock_run = 0;
/*
Modified: head/sys/kern/kern_clocksource.c
==
--- head/sys/kern/kern_clocksource.cMon Nov 22 19:02:30 2010
(r215700)
+++ head/sys/kern/kern_clocksource.cMon Nov 22 19:32:54 2010
(r215701)
@@ -135,7 +135,7 @@ struct pcpu_state {
int idle; /* This CPU is in idle mode. */
};
-STATIC_DPCPU_DEFINE(struct pcpu_state, timerstate);
+static DPCPU_DEFINE(struct pcpu_state, timerstate);
#define FREQ2BT(freq, bt) \
{ \
Modified: head/sys/kern/sched_4bsd.c
==
--- head/sys/kern/sched_4bsd.c Mon Nov 22 19:02:30 2010(r215700)
+++ head/sys/kern/sched_4bsd.c Mon Nov 22 19:32:54 2010(r215701)
@@ -161,7 +161,7 @@ struct pcpuidlestat {
u_int idlecalls;
u_int oldidlecalls;
};
-STATIC_DPCPU_DEFINE(struct pcpuidlestat, idlestat);
+static DPCPU_DEFINE(struct pcpuidlestat, idlestat);
static void
setup_runqs(void)
Modified: head/sys/kern/subr_pcpu.c
==
--- head/sys/kern/subr_pcpu.c Mon Nov 22 19:02:30 2010(r215700)
+++ head/sys/kern/subr_pcpu.c Mon Nov 22
svn commit: r215702 - head/sbin/routed
Author: brucec Date: Mon Nov 22 19:40:27 2010 New Revision: 215702 URL: http://svn.freebsd.org/changeset/base/215702 Log: Fix use of AND operator: should be bitwise instead of logical. Modified: head/sbin/routed/parms.c Modified: head/sbin/routed/parms.c == --- head/sbin/routed/parms.cMon Nov 22 19:32:54 2010(r215701) +++ head/sbin/routed/parms.cMon Nov 22 19:40:27 2010(r215702) @@ -876,11 +876,11 @@ check_parms(struct parm *new) if ((0 != (new-parm_int_state GROUP_IS_SOL_OUT) 0 != (parmp-parm_int_state GROUP_IS_SOL_OUT) 0 != ((new-parm_int_state ^ parmp-parm_int_state) - GROUP_IS_SOL_OUT)) + GROUP_IS_SOL_OUT)) || (0 != (new-parm_int_state GROUP_IS_ADV_OUT) 0 != (parmp-parm_int_state GROUP_IS_ADV_OUT) 0 != ((new-parm_int_state ^ parmp-parm_int_state) - GROUP_IS_ADV_OUT)) + GROUP_IS_ADV_OUT)) || (new-parm_rdisc_pref != 0 parmp-parm_rdisc_pref != 0 new-parm_rdisc_pref != parmp-parm_rdisc_pref) ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215703 - in head/sys: amd64/amd64 i386/i386
Author: jkim
Date: Mon Nov 22 19:52:44 2010
New Revision: 215703
URL: http://svn.freebsd.org/changeset/base/215703
Log:
- Disable caches and flush caches/TLBs when we update PAT as we do for MTRR.
Flushing TLBs is required to ensure cache coherency according to the AMD64
architecture manual. Flushing caches is only required when changing from a
cacheable memory type (WB, WP, or WT) to an uncacheable type (WC, UC, or
UC-). Since this function is only used once per processor during startup,
there is no need to take any shortcuts.
- Leave PAT indices 0-3 at the default of WB, WT, UC-, and UC. Program 5 as
WP (from default WT) and 6 as WC (from default UC-). Leave 4 and 7 at the
default of WB and UC. This is to avoid transition from a cacheable memory
type to an uncacheable type to minimize possible cache incoherency. Since
we perform flushing caches and TLBs now, this change may not be necessary
any more but we do not want to take any chances.
- Remove Apple hardware specific quirks. With the above changes, it seems
this hack is no longer needed.
- Improve pmap_cache_bits() with an array to map PAT memory type to index.
This array is initialized early from pmap_init_pat(), so that we do not need
to handle special cases in the function any more. Now this function is
identical on both amd64 and i386.
Reviewed by: jhb
Tested by:RM (reuf_m at hotmail dot com)
Ryszard Czekaj (rychoo at freeshell dot net)
army.of.root (army dot of dot root at googlemail dot com)
MFC after:3 days
Modified:
head/sys/amd64/amd64/pmap.c
head/sys/i386/i386/pmap.c
Modified: head/sys/amd64/amd64/pmap.c
==
--- head/sys/amd64/amd64/pmap.c Mon Nov 22 19:40:27 2010(r215702)
+++ head/sys/amd64/amd64/pmap.c Mon Nov 22 19:52:44 2010(r215703)
@@ -180,14 +180,20 @@ static vm_paddr_t dmaplimit;
vm_offset_t kernel_vm_end = VM_MIN_KERNEL_ADDRESS;
pt_entry_t pg_nx;
-static int pat_works = 0; /* Is page attribute table sane? */
-
SYSCTL_NODE(_vm, OID_AUTO, pmap, CTLFLAG_RD, 0, VM/pmap parameters);
+static int pat_works = 1;
+TUNABLE_INT(vm.pmap.pat_works, pat_works);
+SYSCTL_INT(_vm_pmap, OID_AUTO, pat_works, CTLFLAG_RDTUN, pat_works, 1,
+Is page attribute table fully functional?);
+
static int pg_ps_enabled = 1;
SYSCTL_INT(_vm_pmap, OID_AUTO, pg_ps_enabled, CTLFLAG_RDTUN, pg_ps_enabled, 0,
Are large page mappings enabled?);
+#definePAT_INDEX_SIZE 8
+static int pat_index[PAT_INDEX_SIZE]; /* cache mode to PAT index conversion */
+
static u_int64_t KPTphys;/* phys addr of kernel level 1 */
static u_int64_t KPDphys;/* phys addr of kernel level 2 */
u_int64_t KPDPphys; /* phys addr of kernel level 3 */
@@ -608,31 +614,24 @@ pmap_bootstrap(vm_paddr_t *firstaddr)
void
pmap_init_pat(void)
{
+ int pat_table[PAT_INDEX_SIZE];
uint64_t pat_msr;
- char *sysenv;
- static int pat_tested = 0;
+ u_long cr0, cr4;
+ int i;
/* Bail if this CPU doesn't implement PAT. */
- if (!(cpu_feature CPUID_PAT))
+ if ((cpu_feature CPUID_PAT) == 0)
panic(no PAT??);
- /*
-* Some Apple Macs based on nVidia chipsets cannot enter ACPI mode
-* via SMI# when we use upper 4 PAT entries for unknown reason.
-*/
- if (!pat_tested) {
- pat_works = 1;
- sysenv = getenv(smbios.system.product);
- if (sysenv != NULL) {
- if (strncmp(sysenv, MacBook5,1, 10) == 0 ||
- strncmp(sysenv, MacBookPro5,5, 13) == 0 ||
- strncmp(sysenv, Macmini3,1, 10) == 0 ||
- strncmp(sysenv, iMac9,1, 7) == 0)
- pat_works = 0;
- freeenv(sysenv);
- }
- pat_tested = 1;
- }
+ /* Set default PAT index table. */
+ for (i = 0; i PAT_INDEX_SIZE; i++)
+ pat_table[i] = -1;
+ pat_table[PAT_WRITE_BACK] = 0;
+ pat_table[PAT_WRITE_THROUGH] = 1;
+ pat_table[PAT_UNCACHEABLE] = 3;
+ pat_table[PAT_WRITE_COMBINING] = 3;
+ pat_table[PAT_WRITE_PROTECTED] = 3;
+ pat_table[PAT_UNCACHED] = 3;
/* Initialize default PAT entries. */
pat_msr = PAT_VALUE(0, PAT_WRITE_BACK) |
@@ -647,20 +646,48 @@ pmap_init_pat(void)
if (pat_works) {
/*
* Leave the indices 0-3 at the default of WB, WT, UC-, and UC.
-* Program 4 and 5 as WP and WC.
-* Leave 6 and 7 as UC- and UC.
+* Program 5 and 6 as WP and WC.
+* Leave 4 and 7 as WB and UC.
*/
- pat_msr = ~(PAT_MASK(4) | PAT_MASK(5));
- pat_msr |= PAT_VALUE(4,
svn commit: r215704 - in head: bin/sh sbin/bsdlabel sbin/geom/class/eli sbin/geom/class/part sbin/geom/class/virstor sbin/growfs sbin/gvinum sbin/hastctl sbin/mknod usr.bin/ar usr.bin/bc
Author: brucec
Date: Mon Nov 22 20:10:48 2010
New Revision: 215704
URL: http://svn.freebsd.org/changeset/base/215704
Log:
Fix some more warnings found by clang.
Modified:
head/bin/sh/arith_lex.l
head/sbin/bsdlabel/bsdlabel.c
head/sbin/geom/class/eli/geom_eli.c
head/sbin/geom/class/part/geom_part.c
head/sbin/geom/class/virstor/geom_virstor.c
head/sbin/growfs/debug.c
head/sbin/gvinum/gvinum.c
head/sbin/hastctl/Makefile
head/sbin/mknod/mknod.c
head/usr.bin/ar/acplex.l
head/usr.bin/bc/scan.l
Modified: head/bin/sh/arith_lex.l
==
--- head/bin/sh/arith_lex.l Mon Nov 22 19:52:44 2010(r215703)
+++ head/bin/sh/arith_lex.l Mon Nov 22 20:10:48 2010(r215704)
@@ -55,6 +55,7 @@ int yylex(void);
#define YY_INPUT(buf,result,max) \
result = (*buf = *arith_buf++) ? 1 : YY_NULL;
#define YY_NO_UNPUT
+#define YY_NO_INPUT
%}
%%
Modified: head/sbin/bsdlabel/bsdlabel.c
==
--- head/sbin/bsdlabel/bsdlabel.c Mon Nov 22 19:52:44 2010
(r215703)
+++ head/sbin/bsdlabel/bsdlabel.c Mon Nov 22 20:10:48 2010
(r215704)
@@ -515,7 +515,7 @@ readlabel(int flag)
f = open(specname, O_RDONLY);
if (f 0)
- err(1, specname);
+ err(1, %s, specname);
if (is_file)
get_file_parms(f);
else {
Modified: head/sbin/geom/class/eli/geom_eli.c
==
--- head/sbin/geom/class/eli/geom_eli.c Mon Nov 22 19:52:44 2010
(r215703)
+++ head/sbin/geom/class/eli/geom_eli.c Mon Nov 22 20:10:48 2010
(r215704)
@@ -374,7 +374,7 @@ eli_genkey_files(struct gctl_req *req, b
if (!gctl_has_param(req, argname))
return (i);
- file = gctl_get_ascii(req, argname);
+ file = gctl_get_ascii(req, %s, argname);
assert(file != NULL);
if (strcmp(file, -) == 0)
Modified: head/sbin/geom/class/part/geom_part.c
==
--- head/sbin/geom/class/part/geom_part.c Mon Nov 22 19:52:44 2010
(r215703)
+++ head/sbin/geom/class/part/geom_part.c Mon Nov 22 20:10:48 2010
(r215704)
@@ -614,7 +614,7 @@ static int
gpart_show_hasopt(struct gctl_req *req, const char *opt, const char *elt)
{
- if (!gctl_get_int(req, opt))
+ if (!gctl_get_int(req, %s, opt))
return (0);
if (elt != NULL)
Modified: head/sbin/geom/class/virstor/geom_virstor.c
==
--- head/sbin/geom/class/virstor/geom_virstor.c Mon Nov 22 19:52:44 2010
(r215703)
+++ head/sbin/geom/class/virstor/geom_virstor.c Mon Nov 22 20:10:48 2010
(r215704)
@@ -276,7 +276,7 @@ virstor_label(struct gctl_req *req)
msize = secsize = 0;
for (i = 1; i (unsigned)nargs; i++) {
snprintf(param, sizeof(param), arg%u, i);
- name = gctl_get_ascii(req, param);
+ name = gctl_get_ascii(req, %s, param);
ssize = g_get_sectorsize(name);
if (ssize == 0)
fprintf(stderr, %s for %s\n, strerror(errno), name);
@@ -336,7 +336,7 @@ virstor_label(struct gctl_req *req)
for (i = 1; i (unsigned)nargs; i++) {
snprintf(param, sizeof(param), arg%u, i);
- name = gctl_get_ascii(req, param);
+ name = gctl_get_ascii(req, %s, param);
if (verbose)
printf( %s, name);
@@ -417,7 +417,7 @@ virstor_label(struct gctl_req *req)
/* Ok, store metadata. */
for (i = 1; i (unsigned)nargs; i++) {
snprintf(param, sizeof(param), arg%u, i);
- name = gctl_get_ascii(req, param);
+ name = gctl_get_ascii(req, %s, param);
msize = g_get_mediasize(name);
ssize = g_get_sectorsize(name);
@@ -434,7 +434,7 @@ virstor_label(struct gctl_req *req)
if (verbose)
printf((%u chunks) , md.chunk_count);
/* Check to make sure last sector is unused */
- if ((off_t)(md.chunk_count * md.md_chunk_size) msize-ssize)
+ if ((off_t)(md.chunk_count * md.md_chunk_size)
(off_t)(msize-ssize))
md.chunk_count--;
md.chunk_next = 0;
if (i != 1) {
@@ -499,7 +499,7 @@ virstor_clear(struct gctl_req *req)
}
for (i = 0; i (unsigned)nargs; i++) {
snprintf(param, sizeof(param), arg%u, i);
- name = gctl_get_ascii(req, param);
+ name = gctl_get_ascii(req, %s, param);
error = g_metadata_clear(name,
svn commit: r215705 - head/usr.bin/ldd
Author: brucec
Date: Mon Nov 22 20:18:46 2010
New Revision: 215705
URL: http://svn.freebsd.org/changeset/base/215705
Log:
hdr.elf.e_ident[EI_OSABI] is not a bitmask so '==' should been used.
Reported by: Artem Belevich fbsdlist at src.cx
Modified:
head/usr.bin/ldd/ldd.c
Modified: head/usr.bin/ldd/ldd.c
==
--- head/usr.bin/ldd/ldd.c Mon Nov 22 20:10:48 2010(r215704)
+++ head/usr.bin/ldd/ldd.c Mon Nov 22 20:18:46 2010(r215705)
@@ -331,7 +331,7 @@ is_executable(const char *fname, int fd,
return (0);
}
if (hdr.elf32.e_type == ET_DYN) {
- if (hdr.elf32.e_ident[EI_OSABI] ELFOSABI_FREEBSD) {
+ if (hdr.elf32.e_ident[EI_OSABI] == ELFOSABI_FREEBSD) {
*is_shlib = 1;
return (1);
}
@@ -373,7 +373,7 @@ is_executable(const char *fname, int fd,
return (0);
}
if (hdr.elf.e_type == ET_DYN) {
- if (hdr.elf.e_ident[EI_OSABI] ELFOSABI_FREEBSD) {
+ if (hdr.elf.e_ident[EI_OSABI] == ELFOSABI_FREEBSD) {
*is_shlib = 1;
return (1);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215706 - head/sys/compat/linux
Author: dim Date: Mon Nov 22 20:23:18 2010 New Revision: 215706 URL: http://svn.freebsd.org/changeset/base/215706 Log: Fix linux kernel module breakage introduced in r215675, by including sys/sysent.h. Noticed by: many Pointy hat to:netchild Modified: head/sys/compat/linux/linux_emul.c Modified: head/sys/compat/linux/linux_emul.c == --- head/sys/compat/linux/linux_emul.c Mon Nov 22 20:18:46 2010 (r215705) +++ head/sys/compat/linux/linux_emul.c Mon Nov 22 20:23:18 2010 (r215706) @@ -41,6 +41,7 @@ __FBSDID($FreeBSD$); #include sys/sx.h #include sys/proc.h #include sys/syscallsubr.h +#include sys/sysent.h #include sys/sysproto.h #include sys/unistd.h ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215707 - head/sys/compat/ndis
Author: bschmidt
Date: Mon Nov 22 20:39:29 2010
New Revision: 215707
URL: http://svn.freebsd.org/changeset/base/215707
Log:
Prefer pmap_extract() over pmap_kextract() as done in MmIsAddressValid().
According to the comment for MmIsAddressValid() there are issues on PAE
kernels using pmap_kextract().
Submitted by: Paul B Mahol onemda at gmail.com
Modified:
head/sys/compat/ndis/subr_ntoskrnl.c
Modified: head/sys/compat/ndis/subr_ntoskrnl.c
==
--- head/sys/compat/ndis/subr_ntoskrnl.cMon Nov 22 20:23:18 2010
(r215706)
+++ head/sys/compat/ndis/subr_ntoskrnl.cMon Nov 22 20:39:29 2010
(r215707)
@@ -2544,6 +2544,12 @@ MmUnmapLockedPages(vaddr, buf)
* here, but it doesn't.
*/
+static uint64_t
+MmGetPhysicalAddress(void *base)
+{
+ return (pmap_extract(kernel_map-pmap, (vm_offset_t)base));
+}
+
uint8_t
MmIsAddressValid(vaddr)
void*vaddr;
@@ -4241,12 +4247,12 @@ image_patch_table ntoskrnl_functbl[] = {
IMPORT_SFUNC(MmAllocateContiguousMemorySpecifyCache, 5 + 3),
IMPORT_SFUNC(MmFreeContiguousMemory, 1),
IMPORT_SFUNC(MmFreeContiguousMemorySpecifyCache, 3),
- IMPORT_SFUNC_MAP(MmGetPhysicalAddress, pmap_kextract, 1),
IMPORT_SFUNC(MmSizeOfMdl, 1),
IMPORT_SFUNC(MmMapLockedPages, 2),
IMPORT_SFUNC(MmMapLockedPagesSpecifyCache, 6),
IMPORT_SFUNC(MmUnmapLockedPages, 2),
IMPORT_SFUNC(MmBuildMdlForNonPagedPool, 1),
+ IMPORT_SFUNC(MmGetPhysicalAddress, 1),
IMPORT_SFUNC(MmIsAddressValid, 1),
IMPORT_SFUNC(MmMapIoSpace, 3 + 1),
IMPORT_SFUNC(MmUnmapIoSpace, 2),
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215708 - head/sys/compat/ndis
Author: bschmidt
Date: Mon Nov 22 20:46:38 2010
New Revision: 215708
URL: http://svn.freebsd.org/changeset/base/215708
Log:
Resurrect amd64 support.
- Many drivers on amd64 are picking system uptime, interrupt time and ticks
via global data structure instead of calling functions for performance
reasons. For now just patch such address so driver will not trigger page
fault when trying to access such data. In future, additional callout may
be added to update data in periodic intervals.
- On amd64 we need to allocate shadow space on stack before calling any
function.
Submitted by: Paul B Mahol onemda at gmail.com
Modified:
head/sys/compat/ndis/kern_windrv.c
head/sys/compat/ndis/ntoskrnl_var.h
head/sys/compat/ndis/subr_ntoskrnl.c
head/sys/compat/ndis/winx64_wrap.S
Modified: head/sys/compat/ndis/kern_windrv.c
==
--- head/sys/compat/ndis/kern_windrv.c Mon Nov 22 20:39:29 2010
(r215707)
+++ head/sys/compat/ndis/kern_windrv.c Mon Nov 22 20:46:38 2010
(r215708)
@@ -311,6 +311,24 @@ windrv_unload(mod, img, len)
#define WINDRV_LOADED htonl(0x42534F44)
+#ifdef __amd64__
+static void
+patch_user_shared_data_address(vm_offset_t img, size_t len)
+{
+ unsigned long i, n, max_addr, *addr;
+
+ n = len - sizeof(unsigned long);
+ max_addr = KI_USER_SHARED_DATA + sizeof(kuser_shared_data);
+ for (i = 0; i n; i++) {
+ addr = (unsigned long *)(img + i);
+ if (*addr = KI_USER_SHARED_DATA *addr max_addr) {
+ *addr -= KI_USER_SHARED_DATA;
+ *addr += (unsigned long)kuser_shared_data;
+ }
+ }
+}
+#endif
+
/*
* Loader routine for actual Windows driver modules, ultimately
* calls the driver's DriverEntry() routine.
@@ -363,6 +381,10 @@ windrv_load(mod, img, len, bustype, devl
return (ENOEXEC);
}
+#ifdef __amd64__
+ patch_user_shared_data_address(img, len);
+#endif
+
/* Dynamically link USBD.SYS -- optional */
if (pe_get_import_descriptor(img, imp_desc, USBD) == 0) {
if (pe_patch_imports(img, USBD, usbd_functbl))
Modified: head/sys/compat/ndis/ntoskrnl_var.h
==
--- head/sys/compat/ndis/ntoskrnl_var.h Mon Nov 22 20:39:29 2010
(r215707)
+++ head/sys/compat/ndis/ntoskrnl_var.h Mon Nov 22 20:46:38 2010
(r215708)
@@ -605,6 +605,65 @@ struct kinterrupt {
typedef struct kinterrupt kinterrupt;
+struct ksystem_time {
+ uint32_tlow_part;
+ int32_t high1_time;
+ int32_t high2_time;
+};
+
+enum nt_product_type {
+ NT_PRODUCT_WIN_NT = 1,
+ NT_PRODUCT_LAN_MAN_NT,
+ NT_PRODUCT_SERVER
+};
+
+enum alt_arch_type {
+ STANDARD_DESIGN,
+ NEC98x86,
+ END_ALTERNATIVES
+};
+
+struct kuser_shared_data {
+ uint32_ttick_count;
+ uint32_ttick_count_multiplier;
+ volatile struct ksystem_time interrupt_time;
+ volatile struct ksystem_time system_time;
+ volatile struct ksystem_time time_zone_bias;
+ uint16_timage_number_low;
+ uint16_timage_number_high;
+ int16_t nt_system_root[260];
+ uint32_tmax_stack_trace_depth;
+ uint32_tcrypto_exponent;
+ uint32_ttime_zone_id;
+ uint32_tlarge_page_min;
+ uint32_treserved2[7];
+ enum nt_product_typent_product_type;
+ uint8_t product_type_is_valid;
+ uint32_tnt_major_version;
+ uint32_tnt_minor_version;
+ uint8_t processor_features[64];
+ uint32_treserved1;
+ uint32_treserved3;
+ volatile uint32_t time_slip;
+ enum alt_arch_type alt_arch_type;
+ int64_t system_expiration_date;
+ uint32_tsuite_mask;
+ uint8_t kdbg_enabled;
+ volatile uint32_t active_console;
+ volatile uint32_t dismount_count;
+ uint32_tcom_plus_package;
+ uint32_tlast_system_rit_event_tick_count;
+ uint32_tnum_phys_pages;
+ uint8_t safe_boot_mode;
+ uint32_ttrace_log;
+ uint64_tfill0;
+ uint64_tsys_call[4];
+ union {
+ volatile struct ksystem_timetick_count;
+ volatile uint64_t tick_count_quad;
+ } tick;
+};
+
/*
* In Windows, there are Physical Device Objects (PDOs) and
* Functional Device Objects (FDOs). Physical Device Objects
svn commit: r215709 - stable/8/sys/vm
Author: mlaier
Date: Mon Nov 22 20:50:40 2010
New Revision: 215709
URL: http://svn.freebsd.org/changeset/base/215709
Log:
MFH r215508:
Off by one page in vm_reserv_reclaim_contig(): reclaim reservations with
only a single free page as well.
Reviewed by:alc
Modified:
stable/8/sys/vm/vm_reserv.c
Directory Properties:
stable/8/sys/ (props changed)
stable/8/sys/amd64/include/xen/ (props changed)
stable/8/sys/cddl/contrib/opensolaris/ (props changed)
stable/8/sys/contrib/dev/acpica/ (props changed)
stable/8/sys/contrib/pf/ (props changed)
stable/8/sys/dev/xen/xenpci/ (props changed)
Modified: stable/8/sys/vm/vm_reserv.c
==
--- stable/8/sys/vm/vm_reserv.c Mon Nov 22 20:46:38 2010(r215708)
+++ stable/8/sys/vm/vm_reserv.c Mon Nov 22 20:50:40 2010(r215709)
@@ -659,7 +659,8 @@ vm_reserv_reclaim_contig(vm_paddr_t size
((pa ^ (pa + size - 1))
~(boundary - 1)) != 0)
pa_length = 0;
- } else if (pa_length = size) {
+ }
+ if (pa_length = size) {
vm_reserv_reclaim(rv);
return (TRUE);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215711 - head/sys/dev/mii
Author: marius
Date: Mon Nov 22 20:57:44 2010
New Revision: 215711
URL: http://svn.freebsd.org/changeset/base/215711
Log:
Add missing newlines.
MFC after:3 days
Modified:
head/sys/dev/mii/mii.c
Modified: head/sys/dev/mii/mii.c
==
--- head/sys/dev/mii/mii.c Mon Nov 22 20:52:18 2010(r215710)
+++ head/sys/dev/mii/mii.c Mon Nov 22 20:57:44 2010(r215711)
@@ -137,7 +137,7 @@ miibus_attach(device_t dev)
free(children, M_TEMP);
}
if (nchildren == 0) {
- device_printf(dev, cannot get children);
+ device_printf(dev, cannot get children\n);
return (ENXIO);
}
ivars = device_get_ivars(dev);
@@ -311,12 +311,12 @@ mii_attach(device_t dev, device_t *miibu
int bmsr, first, i, nchildren, offset, phymax, phymin, rv;
if (phyloc != MII_PHY_ANY offloc != MII_OFFSET_ANY) {
- printf(%s: phyloc and offloc specified, __func__);
+ printf(%s: phyloc and offloc specified\n, __func__);
return (EINVAL);
}
if (offloc != MII_OFFSET_ANY (offloc 0 || offloc = MII_NPHY)) {
- printf(%s: ivalid offloc %d, __func__, offloc);
+ printf(%s: ivalid offloc %d\n, __func__, offloc);
return (EINVAL);
}
@@ -325,7 +325,7 @@ mii_attach(device_t dev, device_t *miibu
phymax = MII_NPHY - 1;
} else {
if (phyloc 0 || phyloc = MII_NPHY) {
- printf(%s: ivalid phyloc %d, __func__, phyloc);
+ printf(%s: ivalid phyloc %d\n, __func__, phyloc);
return (EINVAL);
}
phymin = phymax = phyloc;
@@ -352,7 +352,7 @@ mii_attach(device_t dev, device_t *miibu
if (ivars-ifp != ifp || ivars-ifmedia_upd != ifmedia_upd ||
ivars-ifmedia_sts != ifmedia_sts ||
ivars-mii_flags != flags) {
- printf(%s: non-matching invariant, __func__);
+ printf(%s: non-matching invariant\n, __func__);
return (EINVAL);
}
/*
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215708 - head/sys/compat/ndis
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 11/22/10 12:46, Bernhard Schmidt wrote: Author: bschmidt Date: Mon Nov 22 20:46:38 2010 New Revision: 215708 URL: http://svn.freebsd.org/changeset/base/215708 Log: Resurrect amd64 support. - Many drivers on amd64 are picking system uptime, interrupt time and ticks via global data structure instead of calling functions for performance reasons. For now just patch such address so driver will not trigger page fault when trying to access such data. In future, additional callout may be added to update data in periodic intervals. - On amd64 we need to allocate shadow space on stack before calling any function. Submitted by: Paul B Mahol onemda at gmail.com Will this be MFC'ed? Cheers, - -- Xin LI delp...@delphij.nethttp://www.delphij.net/ FreeBSD - The Power to Serve! Live free or die -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.16 (FreeBSD) iQEcBAEBCAAGBQJM6tuoAAoJEATO+BI/yjfByfcH/15/qNeim1VF5aUK9h6LWPM0 04dB6K3fu3oKEUNY8bD/w06lZZzTPUW/VNrm+bA6/8gKNxP4MgJunty69kmR6tEG xH67RPz/lVmCMU7zNFByFTGltZVSvyyxrsceD6ZylBaSChBV5UoKBMdDE6cOuPbR D5V/yVxlwJaJkhtoRN/GbKlEyZkDs6pGmhzSRKOptr0n4PZjpiHVlXAxhlZkpILB qITYNGvPjyL/3a3nKMz+eBVMqUofjPsDkmCvvApLQumL7FFSCvLGGHnfqe3Pl5Zq nG5ajDia1wnD2nWdbsFoADv69+T3YYhB8wHwvQ/yGV+h+5NExKjovz4l7AZyEjg= =gImO -END PGP SIGNATURE- ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215712 - stable/8/sys/compat/x86bios
Author: jkim Date: Mon Nov 22 21:12:20 2010 New Revision: 215712 URL: http://svn.freebsd.org/changeset/base/215712 Log: Remove a stale file. It seems this file was mis-merged from current. Deleted: stable/8/sys/compat/x86bios/x86bios_alloc.c ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215713 - head/sys/dev/mii
Author: marius Date: Mon Nov 22 21:13:42 2010 New Revision: 215713 URL: http://svn.freebsd.org/changeset/base/215713 Log: Given that unlike f.e. rgephy(4) this driver doesn't explicitly start an autonegotiation along with manual media selection and ukphy_status() also only reports flow control status when BMCR_AUTOEN is set (at least with gentbi(4) determining the flow control status results in false-positives when not set), use MIIF_NOMANPAUSE. Modified: head/sys/dev/mii/ukphy.c Modified: head/sys/dev/mii/ukphy.c == --- head/sys/dev/mii/ukphy.cMon Nov 22 21:12:20 2010(r215712) +++ head/sys/dev/mii/ukphy.cMon Nov 22 21:13:42 2010(r215713) @@ -136,6 +136,8 @@ ukphy_attach(device_t dev) sc-mii_service = ukphy_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + mii_phy_reset(sc); sc-mii_capabilities = PHY_READ(sc, MII_BMSR) ma-mii_capmask; ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215714 - head/sys/dev/mii
Author: marius Date: Mon Nov 22 21:20:11 2010 New Revision: 215714 URL: http://svn.freebsd.org/changeset/base/215714 Log: Given that unlike f.e. rgephy(4) these drivers doen't explicitly start an autonegotiation along with manual media selection and also only report flow control status when BMCR_AUTOEN is set (at least with gentbi(4) determining the flow control status results in false-positives when not set), use MIIF_NOMANPAUSE. Modified: head/sys/dev/mii/gentbi.c head/sys/dev/mii/nsgphy.c Modified: head/sys/dev/mii/gentbi.c == --- head/sys/dev/mii/gentbi.c Mon Nov 22 21:13:42 2010(r215713) +++ head/sys/dev/mii/gentbi.c Mon Nov 22 21:20:11 2010(r215714) @@ -172,6 +172,8 @@ gentbi_attach(device_t dev) sc-mii_service = gentbi_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + mii_phy_reset(sc); /* Modified: head/sys/dev/mii/nsgphy.c == --- head/sys/dev/mii/nsgphy.c Mon Nov 22 21:13:42 2010(r215713) +++ head/sys/dev/mii/nsgphy.c Mon Nov 22 21:20:11 2010(r215714) @@ -135,6 +135,8 @@ nsgphy_attach(device_t dev) sc-mii_service = nsgphy_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + mii_phy_reset(sc); /* ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215715 - stable/7/sys/vm
Author: mlaier
Date: Mon Nov 22 21:22:08 2010
New Revision: 215715
URL: http://svn.freebsd.org/changeset/base/215715
Log:
MFH r215508:
Off by one page in vm_reserv_reclaim_contig(): reclaim reservations with
only a single free page as well.
Reviewed by:alc
Modified:
stable/7/sys/vm/vm_reserv.c
Directory Properties:
stable/7/sys/ (props changed)
stable/7/sys/cddl/contrib/opensolaris/ (props changed)
stable/7/sys/contrib/dev/acpica/ (props changed)
stable/7/sys/contrib/pf/ (props changed)
Modified: stable/7/sys/vm/vm_reserv.c
==
--- stable/7/sys/vm/vm_reserv.c Mon Nov 22 21:20:11 2010(r215714)
+++ stable/7/sys/vm/vm_reserv.c Mon Nov 22 21:22:08 2010(r215715)
@@ -659,7 +659,8 @@ vm_reserv_reclaim_contig(vm_paddr_t size
((pa ^ (pa + size - 1))
~(boundary - 1)) != 0)
pa_length = 0;
- } else if (pa_length = size) {
+ }
+ if (pa_length = size) {
vm_reserv_reclaim(rv);
return (TRUE);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215716 - head/sys/dev/mii
Author: marius Date: Mon Nov 22 21:24:29 2010 New Revision: 215716 URL: http://svn.freebsd.org/changeset/base/215716 Log: Add support for flow control. Obtained from:NetBSD (partially) Modified: head/sys/dev/mii/bmtphy.c head/sys/dev/mii/inphy.c head/sys/dev/mii/nsphyter.c Modified: head/sys/dev/mii/bmtphy.c == --- head/sys/dev/mii/bmtphy.c Mon Nov 22 21:22:08 2010(r215715) +++ head/sys/dev/mii/bmtphy.c Mon Nov 22 21:24:29 2010(r215716) @@ -153,6 +153,8 @@ bmtphy_attach(device_t dev) sc-mii_service = bmtphy_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + mii_phy_reset(sc); sc-mii_capabilities = PHY_READ(sc, MII_BMSR) ma-mii_capmask; @@ -243,7 +245,8 @@ bmtphy_status(struct mii_softc *sc) else mii-mii_media_active |= IFM_10_T; if (aux_csr AUX_CSR_FDX) - mii-mii_media_active |= IFM_FDX; + mii-mii_media_active |= + IFM_FDX | mii_phy_flowstatus(sc); else mii-mii_media_active |= IFM_HDX; } else Modified: head/sys/dev/mii/inphy.c == --- head/sys/dev/mii/inphy.cMon Nov 22 21:22:08 2010(r215715) +++ head/sys/dev/mii/inphy.cMon Nov 22 21:24:29 2010(r215716) @@ -113,6 +113,8 @@ inphy_attach(device_t dev) sc-mii_service = inphy_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + ifmedia_add(mii-mii_media, IFM_MAKEWORD(IFM_ETHER, IFM_100_TX, IFM_LOOP, sc-mii_inst), MII_MEDIA_100_TX, NULL); @@ -197,7 +199,8 @@ inphy_status(struct mii_softc *sc) else mii-mii_media_active |= IFM_10_T; if (scr SCR_FDX) - mii-mii_media_active |= IFM_FDX; + mii-mii_media_active |= + IFM_FDX | mii_phy_flowstatus(sc); else mii-mii_media_active |= IFM_HDX; } else Modified: head/sys/dev/mii/nsphyter.c == --- head/sys/dev/mii/nsphyter.c Mon Nov 22 21:22:08 2010(r215715) +++ head/sys/dev/mii/nsphyter.c Mon Nov 22 21:24:29 2010(r215716) @@ -143,6 +143,8 @@ nsphyter_attach(device_t dev) sc-mii_service = nsphyter_service; sc-mii_pdata = mii; + sc-mii_flags |= MIIF_NOMANPAUSE; + #if 1 #defineADD(m, c) ifmedia_add(mii-mii_media, (m), (c), NULL) @@ -242,12 +244,8 @@ nsphyter_status(struct mii_softc *sc) else mii-mii_media_active |= IFM_100_TX; if ((physts PHYSTS_DUPLEX) != 0) -#ifdef notyet mii-mii_media_active |= IFM_FDX | mii_phy_flowstatus(sc); -#else - mii-mii_media_active |= IFM_FDX; -#endif else mii-mii_media_active |= IFM_HDX; } else ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215708 - head/sys/compat/ndis
On Monday 22 November 2010 22:07:52 Xin LI wrote: On 11/22/10 12:46, Bernhard Schmidt wrote: Author: bschmidt Date: Mon Nov 22 20:46:38 2010 New Revision: 215708 URL: http://svn.freebsd.org/changeset/base/215708 Log: Resurrect amd64 support. - Many drivers on amd64 are picking system uptime, interrupt time and ticks via global data structure instead of calling functions for performance reasons. For now just patch such address so driver will not trigger page fault when trying to access such data. In future, additional callout may be added to update data in periodic intervals. - On amd64 we need to allocate shadow space on stack before calling any function. Submitted by: Paul B Mahol onemda at gmail.com Will this be MFC'ed? I hope I can squeeze it in before the freeze, so yes. Given the current state of ndis on amd64 it definitely won't make things worse. -- Bernhard ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215717 - head/usr.sbin/kernbb
Author: uqs Date: Mon Nov 22 21:39:35 2010 New Revision: 215717 URL: http://svn.freebsd.org/changeset/base/215717 Log: Remove kernbb(8) from the source tree. It's been broken for several years and with all the binutils/toolchain changes in flight, it might make more sense to put efforts into dtrace and hwpmc instead. Discussed with: phk PR: bin/83558 Deleted: head/usr.sbin/kernbb/ ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215719 - head/sys/dev/pci
Author: jkim
Date: Mon Nov 22 21:58:00 2010
New Revision: 215719
URL: http://svn.freebsd.org/changeset/base/215719
Log:
Resume critical PCI devices (and their children) first, then everything else
later. This give us better chance to catch device driver problems.
Modified:
head/sys/dev/pci/pci.c
Modified: head/sys/dev/pci/pci.c
==
--- head/sys/dev/pci/pci.c Mon Nov 22 21:43:45 2010(r215718)
+++ head/sys/dev/pci/pci.c Mon Nov 22 21:58:00 2010(r215719)
@@ -3022,8 +3022,35 @@ pci_resume(device_t dev)
if (!device_is_attached(child))
pci_cfg_save(child, dinfo, 1);
}
+
+ /*
+* Resume critical devices first, then everything else later.
+*/
+ for (i = 0; i numdevs; i++) {
+ child = devlist[i];
+ switch (pci_get_class(child)) {
+ case PCIC_DISPLAY:
+ case PCIC_MEMORY:
+ case PCIC_BRIDGE:
+ case PCIC_BASEPERIPH:
+ DEVICE_RESUME(child);
+ break;
+ }
+ }
+ for (i = 0; i numdevs; i++) {
+ child = devlist[i];
+ switch (pci_get_class(child)) {
+ case PCIC_DISPLAY:
+ case PCIC_MEMORY:
+ case PCIC_BRIDGE:
+ case PCIC_BASEPERIPH:
+ break;
+ default:
+ DEVICE_RESUME(child);
+ }
+ }
free(devlist, M_TEMP);
- return (bus_generic_resume(dev));
+ return (0);
}
static void
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215720 - head/sys/dev/mii
Author: marius
Date: Mon Nov 22 22:03:11 2010
New Revision: 215720
URL: http://svn.freebsd.org/changeset/base/215720
Log:
- Also probe BCM5214 and BCM5222.
- Add some DSP init code for BCM5221. The values derived from Apple's GMAC
driver and the same init code also exists in Linux's sungem_phy driver.
- Only read media status bits when they are valid.
Obtained from:NetBSD, OpenBSD
Modified:
head/sys/dev/mii/bmtphy.c
head/sys/dev/mii/miidevs
Modified: head/sys/dev/mii/bmtphy.c
==
--- head/sys/dev/mii/bmtphy.c Mon Nov 22 21:58:00 2010(r215719)
+++ head/sys/dev/mii/bmtphy.c Mon Nov 22 22:03:11 2010(r215720)
@@ -85,6 +85,11 @@ __FBSDID($FreeBSD$);
static int bmtphy_probe(device_t);
static int bmtphy_attach(device_t);
+struct bmtphy_softc {
+ struct mii_softc mii_sc;
+ int mii_model;
+};
+
static device_method_t bmtphy_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, bmtphy_probe),
@@ -100,18 +105,21 @@ static devclass_t bmtphy_devclass;
static driver_tbmtphy_driver = {
bmtphy,
bmtphy_methods,
- sizeof(struct mii_softc)
+ sizeof(struct bmtphy_softc)
};
DRIVER_MODULE(bmtphy, miibus, bmtphy_driver, bmtphy_devclass, 0, 0);
static int bmtphy_service(struct mii_softc *, struct mii_data *, int);
static voidbmtphy_status(struct mii_softc *);
+static voidbmtphy_reset(struct mii_softc *);
static const struct mii_phydesc bmtphys_dp[] = {
MII_PHY_DESC(BROADCOM, BCM4401),
MII_PHY_DESC(BROADCOM, BCM5201),
+ MII_PHY_DESC(BROADCOM, BCM5214),
MII_PHY_DESC(BROADCOM, BCM5221),
+ MII_PHY_DESC(BROADCOM, BCM5222),
MII_PHY_END
};
@@ -137,11 +145,13 @@ bmtphy_probe(device_t dev)
static int
bmtphy_attach(device_t dev)
{
- struct mii_softc *sc;
- struct mii_attach_args *ma;
- struct mii_data *mii;
+ struct bmtphy_softc *bsc;
+ struct mii_softc *sc;
+ struct mii_attach_args *ma;
+ struct mii_data *mii;
- sc = device_get_softc(dev);
+ bsc = device_get_softc(dev);
+ sc = bsc-mii_sc;
ma = device_get_ivars(dev);
sc-mii_dev = device_get_parent(dev);
mii = ma-mii_data;
@@ -155,7 +165,9 @@ bmtphy_attach(device_t dev)
sc-mii_flags |= MIIF_NOMANPAUSE;
- mii_phy_reset(sc);
+ bsc-mii_model = MII_MODEL(ma-mii_id2);
+
+ bmtphy_reset(sc);
sc-mii_capabilities = PHY_READ(sc, MII_BMSR) ma-mii_capmask;
device_printf(dev, );
@@ -196,16 +208,15 @@ bmtphy_service(struct mii_softc *sc, str
/* Callback if something changed. */
mii_phy_update(sc, cmd);
-
return (0);
}
static void
bmtphy_status(struct mii_softc *sc)
{
- struct mii_data *mii;
- struct ifmedia_entry *ife;
- int bmsr, bmcr, aux_csr;
+ struct mii_data *mii;
+ struct ifmedia_entry *ife;
+ int bmsr, bmcr, aux_csr;
mii = sc-mii_pdata;
ife = mii-mii_media.ifm_cur;
@@ -214,7 +225,6 @@ bmtphy_status(struct mii_softc *sc)
mii-mii_media_active = IFM_ETHER;
bmsr = PHY_READ(sc, MII_BMSR) | PHY_READ(sc, MII_BMSR);
- aux_csr = PHY_READ(sc, MII_BMTPHY_AUX_CSR);
if (bmsr BMSR_LINK)
mii-mii_media_status |= IFM_ACTIVE;
@@ -240,6 +250,7 @@ bmtphy_status(struct mii_softc *sc)
return;
}
+ aux_csr = PHY_READ(sc, MII_BMTPHY_AUX_CSR);
if (aux_csr AUX_CSR_SPEED)
mii-mii_media_active |= IFM_100_TX;
else
@@ -252,3 +263,32 @@ bmtphy_status(struct mii_softc *sc)
} else
mii-mii_media_active = ife-ifm_media;
}
+
+static void
+bmtphy_reset(struct mii_softc *sc)
+{
+ struct bmtphy_softc *bsc;
+ u_int16_t data;
+
+ bsc = (struct bmtphy_softc *)sc;
+
+ mii_phy_reset(sc);
+
+ if (bsc-mii_model == MII_MODEL_BROADCOM_BCM5221) {
+ /* Enable shadow register mode. */
+ data = PHY_READ(sc, 0x1f);
+ PHY_WRITE(sc, 0x1f, data | 0x0080);
+
+ /* Enable APD (Auto PowerDetect). */
+ data = PHY_READ(sc, MII_BMTPHY_AUX2);
+ PHY_WRITE(sc, MII_BMTPHY_AUX2, data | 0x0020);
+
+ /* Enable clocks across APD for Auto-MDIX functionality. */
+ data = PHY_READ(sc, MII_BMTPHY_INTR);
+ PHY_WRITE(sc, MII_BMTPHY_INTR, data | 0x0004);
+
+ /* Disable shadow register mode. */
+ data = PHY_READ(sc, 0x1f);
+ PHY_WRITE(sc, 0x1f, data ~0x0080);
+ }
+}
Modified: head/sys/dev/mii/miidevs
==
--- head/sys/dev/mii/miidevsMon Nov 22 21:58:00 2010(r215719)
+++
Re: svn commit: r215717 - head/usr.sbin/kernbb
On Mon Nov 22 10, Ulrich Spoerlein wrote:
Author: uqs
Date: Mon Nov 22 21:39:35 2010
New Revision: 215717
URL: http://svn.freebsd.org/changeset/base/215717
Log:
Remove kernbb(8) from the source tree.
It's been broken for several years and with all the binutils/toolchain
changes in flight, it might make more sense to put efforts into dtrace and
hwpmc instead.
thanks. :)
what about the BB-profiling sections in
sys/{amd64,i386}/{amd64,i386}/support.S? do they need to stay?
cheers.
alex
Discussed with: phk
PR: bin/83558
Deleted:
head/usr.sbin/kernbb/
--
a13x
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215721 - head/sys/dev/cas
Author: marius
Date: Mon Nov 22 22:06:30 2010
New Revision: 215721
URL: http://svn.freebsd.org/changeset/base/215721
Log:
- Fix and enable support for flow control.
- Fix compilation with CAS_DEBUG defined.
Modified:
head/sys/dev/cas/if_cas.c
Modified: head/sys/dev/cas/if_cas.c
==
--- head/sys/dev/cas/if_cas.c Mon Nov 22 22:03:11 2010(r215720)
+++ head/sys/dev/cas/if_cas.c Mon Nov 22 22:06:30 2010(r215721)
@@ -346,7 +346,7 @@ cas_attach(struct cas_softc *sc)
}
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
cas_mediachange, cas_mediastatus, BMSR_DEFCAPMASK,
- MII_PHY_ANY, MII_OFFSET_ANY, 0);
+ MII_PHY_ANY, MII_OFFSET_ANY, MIIF_DOPAUSE);
}
/*
* Fall back on an internal PHY if no external PHY was found.
@@ -366,7 +366,7 @@ cas_attach(struct cas_softc *sc)
}
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
cas_mediachange, cas_mediastatus, BMSR_DEFCAPMASK,
- MII_PHY_ANY, MII_OFFSET_ANY, 0);
+ MII_PHY_ANY, MII_OFFSET_ANY, MIIF_DOPAUSE);
}
} else {
/*
@@ -388,7 +388,7 @@ cas_attach(struct cas_softc *sc)
BUS_SPACE_BARRIER_READ | BUS_SPACE_BARRIER_WRITE);
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
cas_mediachange, cas_mediastatus, BMSR_DEFCAPMASK,
- CAS_PHYAD_EXTERNAL, MII_OFFSET_ANY, 0);
+ CAS_PHYAD_EXTERNAL, MII_OFFSET_ANY, MIIF_DOPAUSE);
}
if (error != 0) {
device_printf(sc-sc_dev, attaching PHYs failed\n);
@@ -1093,8 +1093,7 @@ cas_init_locked(struct cas_softc *sc)
/* Set the PAUSE thresholds. We use the maximum OFF threshold. */
CAS_WRITE_4(sc, CAS_RX_PTHRS,
- ((111 * 64) CAS_RX_PTHRS_XOFF_SHFT) |
- ((15 * 64) CAS_RX_PTHRS_XON_SHFT));
+ (111 CAS_RX_PTHRS_XOFF_SHFT) | (15 CAS_RX_PTHRS_XON_SHFT));
/* RX blanking */
CAS_WRITE_4(sc, CAS_RX_BLANK,
@@ -1339,7 +1338,7 @@ cas_init_regs(struct cas_softc *sc)
CAS_WRITE_4(sc, CAS_MAC_PREAMBLE_LEN, 0x7);
CAS_WRITE_4(sc, CAS_MAC_JAM_SIZE, 0x4);
CAS_WRITE_4(sc, CAS_MAC_ATTEMPT_LIMIT, 0x10);
- CAS_WRITE_4(sc, CAS_MAC_CTRL_TYPE, 0x8088);
+ CAS_WRITE_4(sc, CAS_MAC_CTRL_TYPE, 0x8808);
/* random number seed */
CAS_WRITE_4(sc, CAS_MAC_RANDOM_SEED,
@@ -1572,11 +1571,11 @@ cas_tint(struct cas_softc *sc)
}
#ifdef CAS_DEBUG
- CTR4(KTR_CAS, %s: CAS_TX_STATE_MACHINE %x CAS_TX_DESC_BASE %llx
+ CTR5(KTR_CAS, %s: CAS_TX_SM1 %x CAS_TX_SM2 %x CAS_TX_DESC_BASE %llx
CAS_TX_COMP3 %x,
- __func__, CAS_READ_4(sc, CAS_TX_STATE_MACHINE),
- ((long long)CAS_READ_4(sc, CAS_TX_DESC_BASE_HI3) 32) |
- CAS_READ_4(sc, CAS_TX_DESC_BASE_LO3),
+ __func__, CAS_READ_4(sc, CAS_TX_SM1), CAS_READ_4(sc, CAS_TX_SM2),
+ ((long long)CAS_READ_4(sc, CAS_TX_DESC3_BASE_HI) 32) |
+ CAS_READ_4(sc, CAS_TX_DESC3_BASE_LO),
CAS_READ_4(sc, CAS_TX_COMP3));
#endif
@@ -1638,7 +1637,7 @@ cas_rint(struct cas_softc *sc)
rxhead = CAS_READ_4(sc, CAS_RX_COMP_HEAD);
#ifdef CAS_DEBUG
CTR4(KTR_CAS, %s: sc-sc_rxcptr %d, sc-sc_rxdptr %d, head %d,
- __func__, sc-rxcptr, sc-sc_rxdptr, rxhead);
+ __func__, sc-sc_rxcptr, sc-sc_rxdptr, rxhead);
#endif
skip = 0;
CAS_CDSYNC(sc, BUS_DMASYNC_POSTREAD | BUS_DMASYNC_POSTWRITE);
@@ -1865,7 +1864,7 @@ cas_rint(struct cas_softc *sc)
#ifdef CAS_DEBUG
CTR4(KTR_CAS, %s: done sc-sc_rxcptr %d, sc-sc_rxdptr %d, head %d,
- __func__, sc-rxcptr, sc-sc_rxdptr,
+ __func__, sc-sc_rxcptr, sc-sc_rxdptr,
CAS_READ_4(sc, CAS_RX_COMP_HEAD));
#endif
}
@@ -1988,7 +1987,7 @@ cas_intr_task(void *arg, int pending __u
#ifdef CAS_DEBUG
CTR4(KTR_CAS, %s: %s: cplt %x, status %x,
device_get_name(sc-sc_dev), __func__,
- (status CAS_STATUS_TX_COMP3_SHIFT), (u_int)status);
+ (status CAS_STATUS_TX_COMP3_SHFT), (u_int)status);
/*
* PCS interrupts must be cleared, otherwise no traffic is passed!
@@ -2100,15 +2099,15 @@ cas_watchdog(struct cas_softc *sc)
#ifdef CAS_DEBUG
CTR4(KTR_CAS,
- %s: CAS_RX_CONFIG %x CAS_MAC_RX_STATUS %x CAS_MAC_RX_CONFIG %x,
- __func__, CAS_READ_4(sc, CAS_RX_CONFIG),
+ %s: CAS_RX_CONF %x CAS_MAC_RX_STATUS %x CAS_MAC_RX_CONF %x,
+ __func__, CAS_READ_4(sc, CAS_RX_CONF),
CAS_READ_4(sc, CAS_MAC_RX_STATUS),
- CAS_READ_4(sc,
svn commit: r215722 - head/sys/dev/gem
Author: marius
Date: Mon Nov 22 22:13:26 2010
New Revision: 215722
URL: http://svn.freebsd.org/changeset/base/215722
Log:
- Fix and enable support for flow control.
- Partially revert r172334; as it turns out the DELAYs in gem_reset_{r,t}x()
are actually necessary although bus space barriers and gem_bitwait() are
used, otherwise the controller may trigger an IOMMU errors on at least
sparc64. This is in line with what Linux and OpenSolaris do.
Modified:
head/sys/dev/gem/if_gem.c
Modified: head/sys/dev/gem/if_gem.c
==
--- head/sys/dev/gem/if_gem.c Mon Nov 22 22:06:30 2010(r215721)
+++ head/sys/dev/gem/if_gem.c Mon Nov 22 22:13:26 2010(r215722)
@@ -302,7 +302,7 @@ gem_attach(struct gem_softc *sc)
}
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
gem_mediachange, gem_mediastatus, BMSR_DEFCAPMASK, phy,
- MII_OFFSET_ANY, 0);
+ MII_OFFSET_ANY, MIIF_DOPAUSE);
}
/*
@@ -330,7 +330,7 @@ gem_attach(struct gem_softc *sc)
}
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
gem_mediachange, gem_mediastatus, BMSR_DEFCAPMASK, phy,
- MII_OFFSET_ANY, 0);
+ MII_OFFSET_ANY, MIIF_DOPAUSE);
}
/*
@@ -352,7 +352,7 @@ gem_attach(struct gem_softc *sc)
sc-sc_flags |= GEM_SERDES;
error = mii_attach(sc-sc_dev, sc-sc_miibus, ifp,
gem_mediachange, gem_mediastatus, BMSR_DEFCAPMASK,
- GEM_PHYAD_EXTERNAL, MII_OFFSET_ANY, 0);
+ GEM_PHYAD_EXTERNAL, MII_OFFSET_ANY, MIIF_DOPAUSE);
}
if (error != 0) {
device_printf(sc-sc_dev, attaching PHYs failed\n);
@@ -712,6 +712,9 @@ gem_reset_rx(struct gem_softc *sc)
if (!GEM_BANK1_BITWAIT(sc, GEM_RX_CONFIG, GEM_RX_CONFIG_RXDMA_EN, 0))
device_printf(sc-sc_dev, cannot disable RX DMA\n);
+ /* Wait 5ms extra. */
+ DELAY(5000);
+
/* Finally, reset the ERX. */
GEM_BANK2_WRITE_4(sc, GEM_RESET, GEM_RESET_RX);
GEM_BANK2_BARRIER(sc, GEM_RESET, 4,
@@ -784,6 +787,9 @@ gem_reset_tx(struct gem_softc *sc)
if (!GEM_BANK1_BITWAIT(sc, GEM_TX_CONFIG, GEM_TX_CONFIG_TXDMA_EN, 0))
device_printf(sc-sc_dev, cannot disable TX DMA\n);
+ /* Wait 5ms extra. */
+ DELAY(5000);
+
/* Finally, reset the ETX. */
GEM_BANK2_WRITE_4(sc, GEM_RESET, GEM_RESET_TX);
GEM_BANK2_BARRIER(sc, GEM_RESET, 4,
@@ -1239,7 +1245,7 @@ gem_init_regs(struct gem_softc *sc)
GEM_BANK1_WRITE_4(sc, GEM_MAC_PREAMBLE_LEN, 0x7);
GEM_BANK1_WRITE_4(sc, GEM_MAC_JAM_SIZE, 0x4);
GEM_BANK1_WRITE_4(sc, GEM_MAC_ATTEMPT_LIMIT, 0x10);
- GEM_BANK1_WRITE_4(sc, GEM_MAC_CONTROL_TYPE, 0x8088);
+ GEM_BANK1_WRITE_4(sc, GEM_MAC_CONTROL_TYPE, 0x8808);
/* random number seed */
GEM_BANK1_WRITE_4(sc, GEM_MAC_RANDOM_SEED,
@@ -2039,14 +2045,12 @@ gem_mii_statchg(device_t dev)
v = GEM_BANK1_READ_4(sc, GEM_MAC_CONTROL_CONFIG)
~(GEM_MAC_CC_RX_PAUSE | GEM_MAC_CC_TX_PAUSE);
-#ifdef notyet
if ((IFM_OPTIONS(sc-sc_mii-mii_media_active)
IFM_ETH_RXPAUSE) != 0)
v |= GEM_MAC_CC_RX_PAUSE;
if ((IFM_OPTIONS(sc-sc_mii-mii_media_active)
IFM_ETH_TXPAUSE) != 0)
v |= GEM_MAC_CC_TX_PAUSE;
-#endif
GEM_BANK1_WRITE_4(sc, GEM_MAC_CONTROL_CONFIG, v);
if ((IFM_OPTIONS(sc-sc_mii-mii_media_active) IFM_FDX) == 0
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r215434 - in head: sys/netinet usr.bin/netstat
On Wed, 17 Nov 2010 18:55:13 + (UTC) George V. Neville-Neil g...@freebsd.org wrote: Author: gnn Date: Wed Nov 17 18:55:12 2010 New Revision: 215434 URL: http://svn.freebsd.org/changeset/base/215434 Log: Add new, per connection, statistics for TCP, including: Retransmitted Packets Zero Window Advertisements Out of Order Receives I don't know if it was this checkin that broke it, but netstat now fails to print a newline between the headers and the first line of connection data. -- Bruce Cran ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215723 - stable/8/sys/netinet/libalias
Author: n_hibma Date: Mon Nov 22 22:41:43 2010 New Revision: 215723 URL: http://svn.freebsd.org/changeset/base/215723 Log: MFC r214754: Don't spam the console with 'Loading alias module' messages. This cannot be hidden behind 'bootverbose' as this file is included from userland. Modified: stable/8/sys/netinet/libalias/alias.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/amd64/include/xen/ (props changed) stable/8/sys/cddl/contrib/opensolaris/ (props changed) stable/8/sys/contrib/dev/acpica/ (props changed) stable/8/sys/contrib/pf/ (props changed) stable/8/sys/dev/xen/xenpci/ (props changed) Modified: stable/8/sys/netinet/libalias/alias.c == --- stable/8/sys/netinet/libalias/alias.c Mon Nov 22 22:13:26 2010 (r215722) +++ stable/8/sys/netinet/libalias/alias.c Mon Nov 22 22:41:43 2010 (r215723) @@ -1665,7 +1665,6 @@ LibAliasRefreshModules(void) if (buf[i] == '#') continue; buf[len - 1] = '\0'; - printf(Loading %s\n, buf); LibAliasLoadModule(buf); } } ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215724 - head/usr.bin/netstat
Author: gnn
Date: Mon Nov 22 22:55:43 2010
New Revision: 215724
URL: http://svn.freebsd.org/changeset/base/215724
Log:
Restore the (state) and \n printout when not using -T.
Pointed out by: brucec@
MFC after:3 weeks
Modified:
head/usr.bin/netstat/inet.c
Modified: head/usr.bin/netstat/inet.c
==
--- head/usr.bin/netstat/inet.c Mon Nov 22 22:41:43 2010(r215723)
+++ head/usr.bin/netstat/inet.c Mon Nov 22 22:55:43 2010(r215724)
@@ -428,13 +428,14 @@ protopr(u_long off, const char *name, in
2msl, delack, rcvtime,
(state));
}
- if (!xflag !Tflag)
+ if (!xflag !Tflag) {
printf((Aflag !Wflag) ?
%-5.5s %-6.6s %-6.6s %-18.18s
%-18.18s :
%-5.5s %-6.6s %-6.6s %-22.22s
%-22.22s,
Proto, Recv-Q, Send-Q,
Local Address, Foreign Address);
-
+ printf((state)\n);
+ }
first = 0;
}
if (Lflag so-so_qlimit == 0)
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215725 - head/sys/dev/ahci
Author: mav Date: Mon Nov 22 23:04:25 2010 New Revision: 215725 URL: http://svn.freebsd.org/changeset/base/215725 Log: Fix small typo. Submitted by: Artem Belevich Modified: head/sys/dev/ahci/ahci.c Modified: head/sys/dev/ahci/ahci.c == --- head/sys/dev/ahci/ahci.cMon Nov 22 22:55:43 2010(r215724) +++ head/sys/dev/ahci/ahci.cMon Nov 22 23:04:25 2010(r215725) @@ -858,7 +858,7 @@ ahci_ch_attach(device_t dev) ch-caps = ctlr-caps; ch-caps2 = ctlr-caps2; ch-quirks = ctlr-quirks; - ch-numslots = ((ch-caps AHCI_CAP_NCS) AHCI_CAP_NCS_SHIFT) + 1, + ch-numslots = ((ch-caps AHCI_CAP_NCS) AHCI_CAP_NCS_SHIFT) + 1; mtx_init(ch-mtx, AHCI channel lock, NULL, MTX_DEF); resource_int_value(device_get_name(dev), device_get_unit(dev), pm_level, ch-pm_level); ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215726 - head/sys/net
Author: zec
Date: Mon Nov 22 23:35:29 2010
New Revision: 215726
URL: http://svn.freebsd.org/changeset/base/215726
Log:
Allow for vlan(4) ifnets to have overlapping unit numbers if they are
created in separated vnets. As a side-effect of having a separated
if_cloner instance for each vnet, all vlan ifnets created in a vnet
will be automatically destroyed when vnet teardown is initiated.
Disallow SIOCSETVLAN and SIOCGETVLAN ioctls on vlan ifnets which are
associated with physical ifnets residing in parent vnets.
This is an interim vlan-specific solution which will be superseded by a
more generic if_cloner V_irtualization change from p4. For nooptions
VIMAGE builds, this should be a no-op change.
Discussed with: bz
MFC after:3 days
Modified:
head/sys/net/if_vlan.c
Modified: head/sys/net/if_vlan.c
==
--- head/sys/net/if_vlan.c Mon Nov 22 23:04:25 2010(r215725)
+++ head/sys/net/if_vlan.c Mon Nov 22 23:35:29 2010(r215726)
@@ -206,6 +206,11 @@ static void vlan_iflladdr(void *arg, st
static struct if_clone vlan_cloner = IFC_CLONE_INITIALIZER(VLANNAME, NULL,
IF_MAXUNIT, NULL, vlan_clone_match, vlan_clone_create, vlan_clone_destroy);
+#ifdef VIMAGE
+static VNET_DEFINE(struct if_clone, vlan_cloner);
+#defineV_vlan_cloner VNET(vlan_cloner)
+#endif
+
#ifndef VLAN_ARRAY
#define HASH(n, m) n) 8) ^ ((n) 4) ^ (n)) (m))
@@ -588,7 +593,9 @@ vlan_modevent(module_t mod, int type, vo
vlan_input_p = vlan_input;
vlan_link_state_p = vlan_link_state;
vlan_trunk_cap_p = vlan_trunk_capabilities;
+#ifndef VIMAGE
if_clone_attach(vlan_cloner);
+#endif
if (bootverbose)
printf(vlan: initialized, using
#ifdef VLAN_ARRAY
@@ -600,7 +607,9 @@ vlan_modevent(module_t mod, int type, vo
\n);
break;
case MOD_UNLOAD:
+#ifndef VIMAGE
if_clone_detach(vlan_cloner);
+#endif
EVENTHANDLER_DEREGISTER(ifnet_departure_event, ifdetach_tag);
EVENTHANDLER_DEREGISTER(iflladdr_event, iflladdr_tag);
vlan_input_p = NULL;
@@ -625,6 +634,27 @@ static moduledata_t vlan_mod = {
DECLARE_MODULE(if_vlan, vlan_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
MODULE_VERSION(if_vlan, 3);
+#ifdef VIMAGE
+static void
+vnet_vlan_init(const void *unused __unused)
+{
+
+ V_vlan_cloner = vlan_cloner;
+ if_clone_attach(V_vlan_cloner);
+}
+VNET_SYSINIT(vnet_vlan_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
+vnet_vlan_init, NULL);
+
+static void
+vnet_vlan_uninit(const void *unused __unused)
+{
+
+ if_clone_detach(V_vlan_cloner);
+}
+VNET_SYSUNINIT(vnet_vlan_uninit, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_FIRST,
+vnet_vlan_uninit, NULL);
+#endif
+
static struct ifnet *
vlan_clone_match_ethertag(struct if_clone *ifc, const char *name, int *tag)
{
@@ -1432,6 +1462,12 @@ vlan_ioctl(struct ifnet *ifp, u_long cmd
break;
case SIOCSETVLAN:
+#ifdef VIMAGE
+ if (ifp-if_vnet != ifp-if_home_vnet) {
+ error = EPERM;
+ break;
+ }
+#endif
error = copyin(ifr-ifr_data, vlr, sizeof(vlr));
if (error)
break;
@@ -1461,6 +1497,12 @@ vlan_ioctl(struct ifnet *ifp, u_long cmd
break;
case SIOCGETVLAN:
+#ifdef VIMAGE
+ if (ifp-if_vnet != ifp-if_home_vnet) {
+ error = EPERM;
+ break;
+ }
+#endif
bzero(vlr, sizeof(vlr));
VLAN_LOCK();
if (TRUNK(ifv) != NULL) {
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215727 - head/bin/sh
Author: jilles
Date: Mon Nov 22 23:49:06 2010
New Revision: 215727
URL: http://svn.freebsd.org/changeset/base/215727
Log:
sh: Fix confusing behaviour if chdir succeeded but getcwd failed in cd -P.
If getcwd fails, do not treat this as an error, but print a warning and
unset PWD. This is similar to the behaviour when starting the shell in a
directory whose name cannot be determined.
Modified:
head/bin/sh/cd.c
Modified: head/bin/sh/cd.c
==
--- head/bin/sh/cd.cMon Nov 22 23:35:29 2010(r215726)
+++ head/bin/sh/cd.cMon Nov 22 23:49:06 2010(r215727)
@@ -219,10 +219,13 @@ cdphysical(char *dest)
char *p;
INTOFF;
- if (chdir(dest) 0 || (p = findcwd(NULL)) == NULL) {
+ if (chdir(dest) 0) {
INTON;
return (-1);
}
+ p = findcwd(NULL);
+ if (p == NULL)
+ out2fmt_flush(cd: warning: failed to get name of current
directory\n);
updatepwd(p);
INTON;
return (0);
@@ -304,7 +307,7 @@ updatepwd(char *dir)
if (prevdir)
ckfree(prevdir);
prevdir = curdir;
- curdir = savestr(dir);
+ curdir = dir ? savestr(dir) : NULL;
setvar(PWD, curdir, VEXPORT);
setvar(OLDPWD, prevdir, VEXPORT);
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215728 - stable/8/sbin/newfs
Author: mckusick
Date: Tue Nov 23 01:24:27 2010
New Revision: 215728
URL: http://svn.freebsd.org/changeset/base/215728
Log:
MFC of 213119
Reported problem:
Large (60GB) filesystems created using newfs -U -O 1 -b 65536 -f 8192
show incorrect results from df for free and used space when mounted
immediately after creation. fsck on the new filesystem (before ever
mounting it once) gives a SUMMARY INFORMATION BAD error in phase 5.
This error hasn't occurred in any runs of fsck immediately after
newfs -U -b 65536 -f 8192 (leaving out the -O 1 option).
Solution:
The default UFS1 superblock is located at offset 8K in the filesystem
partition; the default UFS2 superblock is located at offset 64K in
the filesystem partition. For UFS1 filesystems with a blocksize of
64K, the first alternate superblock resides at 64K which is the the
location used for the default UFS2 superblock. By default, the
system first checks for a valid superblock at the default location
for a UFS2 filoesystem. For a UFS1 filesystem with a blocksize of
64K, there is a valid UFS1 superblock at this location. Thus, even
though it is expected to be a backup superblock, the system will
use it as its default superblock. So, we have to ensure that all the
statistcs on usage are correct in this first alternate superblock
as it is the superblock that will actually be used.
While tracking down this problem, another limitation of UFS1 became
evident. For UFS1, the number of inodes per cylinder group is stored
in an int16_t. Thus the maximum number of inodes per cylinder group
is limited to 2^15 - 1. This limit can easily be exceeded for block
sizes of 32K and above. Thus when building UFS1 filesystems, newfs
must limit the number of inodes per cylinder group to 2^15 - 1.
Reported by: Guy Helmerghel...@palisadesys.com
Followup by: Bruce Cran bru...@freebsd.org
PR: 107692
Modified:
stable/8/sbin/newfs/mkfs.c
Directory Properties:
stable/8/sbin/newfs/ (props changed)
Modified: stable/8/sbin/newfs/mkfs.c
==
--- stable/8/sbin/newfs/mkfs.c Mon Nov 22 23:49:06 2010(r215727)
+++ stable/8/sbin/newfs/mkfs.c Tue Nov 23 01:24:27 2010(r215728)
@@ -376,16 +376,20 @@ restart:
* Start packing more blocks into the cylinder group until
* it cannot grow any larger, the number of cylinder groups
* drops below MINCYLGRPS, or we reach the size requested.
+* For UFS1 inodes per cylinder group are stored in an int16_t
+* so fs_ipg is limited to 2^15 - 1.
*/
for ( ; sblock.fs_fpg maxblkspercg; sblock.fs_fpg += sblock.fs_frag) {
sblock.fs_ipg = roundup(howmany(sblock.fs_fpg, fragsperinode),
INOPB(sblock));
- if (sblock.fs_size / sblock.fs_fpg MINCYLGRPS)
- break;
- if (CGSIZE(sblock) (unsigned long)sblock.fs_bsize)
- continue;
- if (CGSIZE(sblock) == (unsigned long)sblock.fs_bsize)
- break;
+ if (Oflag 1 || (Oflag == 1 sblock.fs_ipg = 0x7fff)) {
+ if (sblock.fs_size / sblock.fs_fpg MINCYLGRPS)
+ break;
+ if (CGSIZE(sblock) (unsigned long)sblock.fs_bsize)
+ continue;
+ if (CGSIZE(sblock) == (unsigned long)sblock.fs_bsize)
+ break;
+ }
sblock.fs_fpg -= sblock.fs_frag;
sblock.fs_ipg = roundup(howmany(sblock.fs_fpg, fragsperinode),
INOPB(sblock));
@@ -584,8 +588,20 @@ restart:
printf(** Exiting on Xflag 3\n);
exit(0);
}
- if (!Nflag)
+ if (!Nflag) {
do_sbwrite(disk);
+ /*
+* For UFS1 filesystems with a blocksize of 64K, the first
+* alternate superblock resides at the location used for
+* the default UFS2 superblock. As there is a valid
+* superblock at this location, the boot code will use
+* it as its first choice. Thus we have to ensure that
+* all of its statistcs on usage are correct.
+*/
+ if (Oflag == 1 sblock.fs_bsize == 65536)
+ wtfs(fsbtodb(sblock, cgsblock(sblock, 0)),
+ sblock.fs_bsize, (char *)sblock);
+ }
for (i = 0; i sblock.fs_cssize; i += sblock.fs_bsize)
wtfs(fsbtodb(sblock, sblock.fs_csaddr + numfrags(sblock, i)),
sblock.fs_cssize - i sblock.fs_bsize ?
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to
svn commit: r215729 - stable/7/sbin/newfs
Author: mckusick
Date: Tue Nov 23 01:32:44 2010
New Revision: 215729
URL: http://svn.freebsd.org/changeset/base/215729
Log:
MFC of 213119
Reported problem:
Large (60GB) filesystems created using newfs -U -O 1 -b 65536 -f 8192
show incorrect results from df for free and used space when mounted
immediately after creation. fsck on the new filesystem (before ever
mounting it once) gives a SUMMARY INFORMATION BAD error in phase 5.
This error hasn't occurred in any runs of fsck immediately after
newfs -U -b 65536 -f 8192 (leaving out the -O 1 option).
Solution:
The default UFS1 superblock is located at offset 8K in the filesystem
partition; the default UFS2 superblock is located at offset 64K in
the filesystem partition. For UFS1 filesystems with a blocksize of
64K, the first alternate superblock resides at 64K which is the the
location used for the default UFS2 superblock. By default, the
system first checks for a valid superblock at the default location
for a UFS2 filoesystem. For a UFS1 filesystem with a blocksize of
64K, there is a valid UFS1 superblock at this location. Thus, even
though it is expected to be a backup superblock, the system will
use it as its default superblock. So, we have to ensure that all the
statistcs on usage are correct in this first alternate superblock
as it is the superblock that will actually be used.
While tracking down this problem, another limitation of UFS1 became
evident. For UFS1, the number of inodes per cylinder group is stored
in an int16_t. Thus the maximum number of inodes per cylinder group
is limited to 2^15 - 1. This limit can easily be exceeded for block
sizes of 32K and above. Thus when building UFS1 filesystems, newfs
must limit the number of inodes per cylinder group to 2^15 - 1.
Reported by: Guy Helmerghel...@palisadesys.com
Followup by: Bruce Cran bru...@freebsd.org
PR: 107692
Modified:
stable/7/sbin/newfs/mkfs.c
Directory Properties:
stable/7/sbin/newfs/ (props changed)
Modified: stable/7/sbin/newfs/mkfs.c
==
--- stable/7/sbin/newfs/mkfs.c Tue Nov 23 01:24:27 2010(r215728)
+++ stable/7/sbin/newfs/mkfs.c Tue Nov 23 01:32:44 2010(r215729)
@@ -367,16 +367,20 @@ restart:
* Start packing more blocks into the cylinder group until
* it cannot grow any larger, the number of cylinder groups
* drops below MINCYLGRPS, or we reach the size requested.
+* For UFS1 inodes per cylinder group are stored in an int16_t
+* so fs_ipg is limited to 2^15 - 1.
*/
for ( ; sblock.fs_fpg maxblkspercg; sblock.fs_fpg += sblock.fs_frag) {
sblock.fs_ipg = roundup(howmany(sblock.fs_fpg, fragsperinode),
INOPB(sblock));
- if (sblock.fs_size / sblock.fs_fpg MINCYLGRPS)
- break;
- if (CGSIZE(sblock) (unsigned long)sblock.fs_bsize)
- continue;
- if (CGSIZE(sblock) == (unsigned long)sblock.fs_bsize)
- break;
+ if (Oflag 1 || (Oflag == 1 sblock.fs_ipg = 0x7fff)) {
+ if (sblock.fs_size / sblock.fs_fpg MINCYLGRPS)
+ break;
+ if (CGSIZE(sblock) (unsigned long)sblock.fs_bsize)
+ continue;
+ if (CGSIZE(sblock) == (unsigned long)sblock.fs_bsize)
+ break;
+ }
sblock.fs_fpg -= sblock.fs_frag;
sblock.fs_ipg = roundup(howmany(sblock.fs_fpg, fragsperinode),
INOPB(sblock));
@@ -568,8 +572,20 @@ restart:
printf(** Exiting on Eflag 3\n);
exit(0);
}
- if (!Nflag)
+ if (!Nflag) {
sbwrite(disk, 0);
+ /*
+* For UFS1 filesystems with a blocksize of 64K, the first
+* alternate superblock resides at the location used for
+* the default UFS2 superblock. As there is a valid
+* superblock at this location, the boot code will use
+* it as its first choice. Thus we have to ensure that
+* all of its statistcs on usage are correct.
+*/
+ if (Oflag == 1 sblock.fs_bsize == 65536)
+ wtfs(fsbtodb(sblock, cgsblock(sblock, 0)),
+ sblock.fs_bsize, (char *)sblock);
+ }
for (i = 0; i sblock.fs_cssize; i += sblock.fs_bsize)
wtfs(fsbtodb(sblock, sblock.fs_csaddr + numfrags(sblock, i)),
sblock.fs_cssize - i sblock.fs_bsize ?
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to
svn commit: r215730 - in stable/8: etc/periodic/security tools/build/mk
Author: delphij
Date: Tue Nov 23 01:39:30 2010
New Revision: 215730
URL: http://svn.freebsd.org/changeset/base/215730
Log:
MFC r204459 (antonie):
Add files to remove when MK_PKGTOOLS=no.
MFC r215045 [1]:
Hide 460.chkportsum in MK_PKGTOOLS != no case.
Submitted by: Alex Kozlov spam rm-rf kiev ua [1]
Modified:
stable/8/etc/periodic/security/Makefile
stable/8/tools/build/mk/OptionalObsoleteFiles.inc
Directory Properties:
stable/8/etc/periodic/security/ (props changed)
stable/8/tools/build/mk/ (props changed)
Modified: stable/8/etc/periodic/security/Makefile
==
--- stable/8/etc/periodic/security/Makefile Tue Nov 23 01:32:44 2010
(r215729)
+++ stable/8/etc/periodic/security/Makefile Tue Nov 23 01:39:30 2010
(r215730)
@@ -7,7 +7,6 @@ FILES= 100.chksetuid \
300.chkuid0 \
400.passwdless \
410.logincheck \
- 460.chkportsum \
700.kernelmsg \
800.loginfail \
900.tcpwrap \
@@ -28,4 +27,8 @@ FILES+= 500.ipfwdenied \
FILES+=520.pfdenied
.endif
+.if ${MK_PKGTOOLS} != no
+FILES+=460.chkportsum
+.endif
+
.include bsd.prog.mk
Modified: stable/8/tools/build/mk/OptionalObsoleteFiles.inc
==
--- stable/8/tools/build/mk/OptionalObsoleteFiles.inc Tue Nov 23 01:32:44
2010(r215729)
+++ stable/8/tools/build/mk/OptionalObsoleteFiles.inc Tue Nov 23 01:39:30
2010(r215730)
@@ -1213,6 +1213,23 @@ OLD_FILES+=usr/share/man/man1/nc.1.gz
# to be filled in
#.endif
+.if ${MK_PKGTOOLS} == no
+OLD_FILES+=etc/periodic/security/460.chkportsum
+OLD_FILES+=etc/periodic/weekly/400.status-pkg
+OLD_FILES+=usr/sbin/pkg_add
+OLD_FILES+=usr/sbin/pkg_create
+OLD_FILES+=usr/sbin/pkg_delete
+OLD_FILES+=usr/sbin/pkg_info
+OLD_FILES+=usr/sbin/pkg_updating
+OLD_FILES+=usr/sbin/pkg_version
+OLD_FILES+=usr/share/man/man1/pkg_add.1.gz
+OLD_FILES+=usr/share/man/man1/pkg_create.1.gz
+OLD_FILES+=usr/share/man/man1/pkg_delete.1.gz
+OLD_FILES+=usr/share/man/man1/pkg_info.1.gz
+OLD_FILES+=usr/share/man/man1/pkg_updating.1.gz
+OLD_FILES+=usr/share/man/man1/pkg_version.1.gz
+.endif
+
.if ${MK_PROFILE} == no
OLD_FILES+=usr/lib/libalias_cuseeme_p.a
OLD_FILES+=usr/lib/libalias_dummy_p.a
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215732 - head/sys/kern
Author: cperciva
Date: Tue Nov 23 04:50:01 2010
New Revision: 215732
URL: http://svn.freebsd.org/changeset/base/215732
Log:
Add parentheses for clarity. The parentheses around the two terms of the
are unnecessary but I'm leaving them in for the sake of avoiding confusion
(I confuse easily).
Submitted by: bde
Modified:
head/sys/kern/kern_tc.c
Modified: head/sys/kern/kern_tc.c
==
--- head/sys/kern/kern_tc.c Tue Nov 23 03:53:53 2010(r215731)
+++ head/sys/kern/kern_tc.c Tue Nov 23 04:50:01 2010(r215732)
@@ -448,7 +448,7 @@ tc_windup(void)
th-th_offset.sec++;
}
if ((delta th-th_counter-tc_frequency / 2)
- (th-th_scale * delta (uint64_t)1 63)) {
+ (th-th_scale * delta ((uint64_t)1 63))) {
/* The product th_scale * delta just barely overflows. */
th-th_offset.sec++;
}
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r215733 - head/share/misc
Author: maxim Date: Tue Nov 23 06:31:09 2010 New Revision: 215733 URL: http://svn.freebsd.org/changeset/base/215733 Log: o NetBSD 5.1 added. Modified: head/share/misc/bsd-family-tree Modified: head/share/misc/bsd-family-tree == --- head/share/misc/bsd-family-tree Tue Nov 23 04:50:01 2010 (r215732) +++ head/share/misc/bsd-family-tree Tue Nov 23 06:31:09 2010 (r215733) @@ -228,20 +228,20 @@ FreeBSD 5.2 | | | FreeBSD 7.1 | | | | | | | | |DragonFly 2.2.0 | FreeBSD 7.2 | NetBSD 5.0 OpenBSD 4.5 | - | \ | | | | - | | | | |DragonFly 2.4.0 - | | | | OpenBSD 4.6 | - | | | | | | - *--FreeBSD | | | | | - |8.0 | | | | | - | |FreeBSD | | | | - | | 7.3| | |DragonFly 2.6.0 - | | | | OpenBSD 4.7 | - | FreeBSD | | | | - |8.1 | | | | - | | | | |DragonFly 2.8.0 - | | | | OpenBSD 4.8 | - | V | | | | + | \ | ||| | + | | | |||DragonFly 2.4.0 + | | | ||OpenBSD 4.6 | + | | | ||| | + *--FreeBSD | | ||| | + |8.0 | | ||| | + | |FreeBSD | ||| | + | | 7.3| |||DragonFly 2.6.0 + | | | ||OpenBSD 4.7 | + | FreeBSD | ||| | + |8.1 | ||| | + | | | |||DragonFly 2.8.0 + | | | ||OpenBSD 4.8 | + | V | | NetBSD 5.1 | | || | | | FreeBSD 9 -current| NetBSD -current OpenBSD -current | || | | | @@ -523,6 +523,7 @@ OpenBSD 4.7 2010-05-19 [OBD] FreeBSD 8.12010-07-24 [FBD] DragonFly 2.8.02010-10-30 [DFB] OpenBSD 4.82010-11-01 [OBD] +NetBSD 5.1 2010-11-19 [NBD] Bibliography ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
