Re: svn commit: r213744 - head/bin/sh
On Wed, 13 Oct 2010, John Baldwin wrote: On Wednesday, October 13, 2010 12:29:27 am Bruce Evans wrote: On Tue, 12 Oct 2010, David O'Brien wrote: On Wed, Oct 13, 2010 at 02:18:33PM +1100, Bruce Evans wrote: ... Of course, debugging and profiling are magic, but I don't want to have to adorn all functions with STATICs and __attributes() (and pragmas for othercc...) to recover historical/normal or variant debugging or profiling of them. I agree, and would not add STATIC's to a program's code that didn't already have them. But in this case we inherited it from 4.4BSD. I'm just making it actually do something other than being a gratuitous spelling change. I believe I've made things more consistent with r213760. Add __noinline or whatever attributes to STATIC (but keep static in it) for the DEBUG >= 3 case if you are going that far. __noinline should be a syntax error for variables, so this should also find any STATICs still on variables. The spelling fix of changing STATIC to what it actually means (ASSORTED_HACKS_FOR_DEBUGGING_BUT_NOW_ONLY_FOR_FUNCTIONS) goes too far for me. To be honest, I think changing STATIC is excessive churn. I agree, but 'static' was already spelled 'STATIC' in all cases without style bugs (except in the #define of STATIC). Adding __noinline to the #define of STATIC is a small change. The "right" fix is to use 'make DEBUG_FLAGS="-g -DDEBUG=2 -fno-inline"'. I often use 'make DEBUG_FLAGS="-g -fno-inline"' to workaround excessive inlining when debugging. I think all of the STATIC changes should just be backed out. Adding -fno-inline on the command line is more flexible, but might change the generated code too much. Sometimes you want to keep all the little inline functions inline so as not to see them, or just to give the same code for them. -fno-inline-functions-called-once may be more useful. -fno-inline doesn't affect functions declared as __inline, so in the kernel where __inline is used too much, -fno-inline doesn't have such a large effect. Putting the non-inlining flag in the definition of STATIC gives even finer control. But I would only use it if STATIC is already used. I would try the hack of "#define static /**/" or "#define static __noinline_mumble" before changing the source code. Bruce ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213793 - in head/sys/dev: ce cp
On Wed, 13 Oct 2010, Jung-uk Kim wrote:
On Wednesday 13 October 2010 01:27 pm, Roman Divacky wrote:
Modified: head/sys/dev/ce/if_ce.c
=
= --- head/sys/dev/ce/if_ce.c Wed Oct 13 17:16:08
2010(r213792) +++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:17:50
2010(r213793) @@ -1313,7 +1313,7 @@ static int ce_ioctl (struct
cdev *dev, u IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data) && PP_FR) {
+ } else if (! strcmp ("fr", (char*)data)) {
this is wrong I think... the PP_FR was used for compiling in/out
support for something.. see the comment:
/* If we don't have Cronyx's sppp version, we don't have fr support
via sppp */ #ifndef PP_FR
#define PP_FR 0
#endif
note that PP_FR is used in some other places as a flag. I guess
that by compiling with something like make -DPP_FR=42 some magic
would happen.
anyway - this does not look like a bug but like an intent, please
revert.
I think the attached patch should do.
% Index: sys/dev/ce/if_ce.c
% ===
% --- sys/dev/ce/if_ce.c(revision 213782)
% +++ sys/dev/ce/if_ce.c(working copy)
% @@ -1313,9 +1313,11 @@ static int ce_ioctl (struct cdev *dev, u_long cmd,
% IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
% IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
% d->ifp->if_flags |= PP_CISCO;
% - } else if (! strcmp ("fr", (char*)data) && PP_FR) {
% +#if PP_FR > 0
% + } else if (! strcmp ("fr", (char*)data)) {
% d->ifp->if_flags &= ~(PP_CISCO);
% IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
% +#endif
% } else if (! strcmp ("ppp", (char*)data)) {
% IFP2SP(d->ifp)->pp_flags &= ~PP_FR;
% IFP2SP(d->ifp)->pp_flags &= ~PP_KEEPALIVE;
% ...
This gives different behaviour if PP_FR is even or negative. Even values
used to fail the match, but now the match succeeds for even values > 0 and
then the bits of PP_FR are put in pp_flags. Negative values used to pass
the match if they were odd, but now the match is not attempted for any
negative value. It just might be useful for PP_FR to have multiple bits
set, with the 1 bit used for enabling this and the other bits used for
setting pp_flags. If not, then only values of 0 and 1 for PP_FR make sense,
and the ifdef should be "#if PP_FR == 1".
One of the 3 style bugs of "! strcmp (...)" is larger now. The normal
style of "strcmp(...) == 0" would have inhibited this bug, but perhaps
"!" was used to emphasize that the result is boolean. The 3 strcmp's
visible in the above patch are the only ones in the file with the "!
" style bugs. The normal style of explicitly comparing the result of
strcmp with 0 is used 2 times, with consistency only for using a
gnu-style space before the left parentheses.
I don't see why clang complained about this. "! strcmp()" has a boolean
result with value of 1 for "true". It is perfectly valid to AND
this result with a boolean flag that also has value 1 for "true",
or with an integral bitmap that uses bit 0 (value 1) for this test.
Bruce
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213815 - head/bin/sh
Author: obrien
Date: Wed Oct 13 23:31:17 2010
New Revision: 213815
URL: http://svn.freebsd.org/changeset/base/213815
Log:
We only need to look as far as '..' to find 'test/'.
Modified:
head/bin/sh/Makefile
Modified: head/bin/sh/Makefile
==
--- head/bin/sh/MakefileWed Oct 13 23:29:09 2010(r213814)
+++ head/bin/sh/MakefileWed Oct 13 23:31:17 2010(r213815)
@@ -26,7 +26,7 @@ WARNS?= 2
WFORMAT=0
.PATH: ${.CURDIR}/bltin \
- ${.CURDIR}/../../bin/test
+ ${.CURDIR}/../test
CLEANFILES+= mkinit mkinit.o mknodes mknodes.o \
mksyntax mksyntax.o
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213814 - in head: bin/sh tools/regression/bin/sh/expansion
Author: obrien
Date: Wed Oct 13 23:29:09 2010
New Revision: 213814
URL: http://svn.freebsd.org/changeset/base/213814
Log:
Do not assume in growstackstr() that a "precious" character will be
immediately written into the stack after the call. Instead let the caller
manage the "space left".
Previously, growstackstr()'s assumption causes problems with STACKSTRNUL()
where we want to be able to turn a stack into a C string, and later
pretend the NUL is not there.
This fixes a bug in STACKSTRNUL() (that grew the stack) where:
1. STADJUST() called after a STACKSTRNUL() results in an improper adjust.
This can be seen in ${var%pattern} and ${var%%pattern} evaluation.
2. Memory leak in STPUTC() called after a STACKSTRNUL().
Reviewed by: jilles
Added:
head/tools/regression/bin/sh/expansion/trim4.0 (contents, props changed)
Modified:
head/bin/sh/histedit.c
head/bin/sh/memalloc.c
head/bin/sh/memalloc.h
Modified: head/bin/sh/histedit.c
==
--- head/bin/sh/histedit.c Wed Oct 13 22:59:04 2010(r213813)
+++ head/bin/sh/histedit.c Wed Oct 13 23:29:09 2010(r213814)
@@ -418,7 +418,7 @@ fc_replace(const char *s, char *p, char
} else
STPUTC(*s++, dest);
}
- STACKSTRNUL(dest);
+ STPUTC('\0', dest);
dest = grabstackstr(dest);
return (dest);
Modified: head/bin/sh/memalloc.c
==
--- head/bin/sh/memalloc.c Wed Oct 13 22:59:04 2010(r213813)
+++ head/bin/sh/memalloc.c Wed Oct 13 23:29:09 2010(r213814)
@@ -295,6 +295,13 @@ grabstackblock(int len)
* is space for at least one character.
*/
+static char *
+growstrstackblock(int n)
+{
+ growstackblock();
+ sstrnleft = stackblocksize() - n;
+ return stackblock() + n;
+}
char *
growstackstr(void)
@@ -304,12 +311,10 @@ growstackstr(void)
len = stackblocksize();
if (herefd >= 0 && len >= 1024) {
xwrite(herefd, stackblock(), len);
- sstrnleft = len - 1;
+ sstrnleft = len;
return stackblock();
}
- growstackblock();
- sstrnleft = stackblocksize() - len - 1;
- return stackblock() + len;
+ return growstrstackblock(len);
}
@@ -323,9 +328,7 @@ makestrspace(void)
int len;
len = stackblocksize() - sstrnleft;
- growstackblock();
- sstrnleft = stackblocksize() - len;
- return stackblock() + len;
+ return growstrstackblock(len);
}
Modified: head/bin/sh/memalloc.h
==
--- head/bin/sh/memalloc.h Wed Oct 13 22:59:04 2010(r213813)
+++ head/bin/sh/memalloc.h Wed Oct 13 23:29:09 2010(r213814)
@@ -67,9 +67,16 @@ void ungrabstackstr(char *, char *);
#define stackblock() stacknxt
#define stackblocksize() stacknleft
#define STARTSTACKSTR(p) p = stackblock(), sstrnleft = stackblocksize()
-#define STPUTC(c, p) (--sstrnleft >= 0? (*p++ = (c)) : (p = growstackstr(),
*p++ = (c)))
+#define STPUTC(c, p) (--sstrnleft >= 0? (*p++ = (c)) : (p = growstackstr(),
--sstrnleft, *p++ = (c)))
#define CHECKSTRSPACE(n, p){ if (sstrnleft < n) p = makestrspace(); }
#define USTPUTC(c, p) (--sstrnleft, *p++ = (c))
+/*
+ * STACKSTRNUL's use is where we want to be able to turn a stack
+ * (non-sentinel, character counting string) into a C string,
+ * and later pretend the NUL is not there.
+ * Note: Because of STACKSTRNUL's semantics, STACKSTRNUL cannot be used
+ * on a stack that will grabstackstr()ed.
+ */
#define STACKSTRNUL(p) (sstrnleft == 0? (p = growstackstr(), *p = '\0') : (*p
= '\0'))
#define STUNPUTC(p)(++sstrnleft, --p)
#define STTOPC(p) p[-1]
Added: head/tools/regression/bin/sh/expansion/trim4.0
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/tools/regression/bin/sh/expansion/trim4.0 Wed Oct 13 23:29:09
2010(r213814)
@@ -0,0 +1,15 @@
+# $FreeBSD$
+
+v1=/homes/SOME_USER
+v2=
+v3=C123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789
+
+# Trigger bug in VSTRIMRIGHT processing STADJUST() call in
expand.c:subevalvar()
+while [ ${#v2} -lt 2000 ]; do
+ v4="${v2} ${v1%/*} $v3"
+ if [ ${#v4} -ne $((${#v2} + ${#v3} + 8)) ]; then
+ echo bad: ${#v4} -ne $((${#v2} + ${#v3} + 8))
+ fi
+ v2=x$v2
+ v3=y$v3
+done
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213813 - in head: share/man/man9 sys/kern sys/sys
Author: mdf
Date: Wed Oct 13 22:59:04 2010
New Revision: 213813
URL: http://svn.freebsd.org/changeset/base/213813
Log:
Use a safer mechanism for determining if a task is currently running,
that does not rely on the lifetime of pointers being the same. This also
restores the task KBI.
Suggested by: jhb
MFC after:1 month
Modified:
head/share/man/man9/taskqueue.9
head/sys/kern/subr_taskqueue.c
head/sys/sys/_task.h
head/sys/sys/taskqueue.h
Modified: head/share/man/man9/taskqueue.9
==
--- head/share/man/man9/taskqueue.9 Wed Oct 13 22:29:48 2010
(r213812)
+++ head/share/man/man9/taskqueue.9 Wed Oct 13 22:59:04 2010
(r213813)
@@ -68,7 +68,7 @@ struct task {
.Ft int
.Fn taskqueue_member "struct taskqueue *queue" "struct thread *td"
.Ft void
-.Fn taskqueue_run "struct taskqueue *queue" "struct task **tpp"
+.Fn taskqueue_run "struct taskqueue *queue"
.Fn TASK_INIT "struct task *task" "int priority" "task_fn_t *func" "void
*context"
.Fn TASKQUEUE_DECLARE "name"
.Fn TASKQUEUE_DEFINE "name" "taskqueue_enqueue_fn enqueue" "void *context"
"init"
@@ -185,13 +185,6 @@ The
function will run all pending tasks in the specified
.Fa queue .
Normally this function is only used internally.
-The
-.Fa tpp
-argument is a pointer to a
-.Vt struct task *
-that is used to record the currently running task.
-The lifetime of this pointer must match that of the
-.Fa queue .
.Pp
A convenience macro,
.Fn TASK_INIT "task" "priority" "func" "context"
Modified: head/sys/kern/subr_taskqueue.c
==
--- head/sys/kern/subr_taskqueue.c Wed Oct 13 22:29:48 2010
(r213812)
+++ head/sys/kern/subr_taskqueue.c Wed Oct 13 22:59:04 2010
(r213813)
@@ -46,12 +46,17 @@ static MALLOC_DEFINE(M_TASKQUEUE, "taskq
static void*taskqueue_giant_ih;
static void*taskqueue_ih;
+struct taskqueue_busy {
+ struct task *tb_running;
+ TAILQ_ENTRY(taskqueue_busy) tb_link;
+};
+
struct taskqueue {
STAILQ_HEAD(, task) tq_queue;
const char *tq_name;
taskqueue_enqueue_fntq_enqueue;
void*tq_context;
- struct task *tq_running;
+ TAILQ_HEAD(, taskqueue_busy) tq_active;
struct mtx tq_mutex;
struct thread **tq_threads;
int tq_tcount;
@@ -102,6 +107,7 @@ _taskqueue_create(const char *name, int
return NULL;
STAILQ_INIT(&queue->tq_queue);
+ TAILQ_INIT(&queue->tq_active);
queue->tq_name = name;
queue->tq_enqueue = enqueue;
queue->tq_context = context;
@@ -140,6 +146,7 @@ taskqueue_free(struct taskqueue *queue)
TQ_LOCK(queue);
queue->tq_flags &= ~TQ_FLAGS_ACTIVE;
taskqueue_terminate(queue->tq_threads, queue);
+ KASSERT(TAILQ_EMPTY(&queue->tq_active), ("Tasks still running?"));
mtx_destroy(&queue->tq_mutex);
free(queue->tq_threads, M_TASKQUEUE);
free(queue, M_TASKQUEUE);
@@ -214,13 +221,17 @@ taskqueue_unblock(struct taskqueue *queu
TQ_UNLOCK(queue);
}
-void
-taskqueue_run(struct taskqueue *queue, struct task **tpp)
+static void
+taskqueue_run_locked(struct taskqueue *queue)
{
+ struct taskqueue_busy tb;
struct task *task;
int pending;
mtx_assert(&queue->tq_mutex, MA_OWNED);
+ tb.tb_running = NULL;
+ TAILQ_INSERT_TAIL(&queue->tq_active, &tb, tb_link);
+
while (STAILQ_FIRST(&queue->tq_queue)) {
/*
* Carefully remove the first task from the queue and
@@ -230,16 +241,38 @@ taskqueue_run(struct taskqueue *queue, s
STAILQ_REMOVE_HEAD(&queue->tq_queue, ta_link);
pending = task->ta_pending;
task->ta_pending = 0;
- task->ta_running = tpp;
- *tpp = task;
+ tb.tb_running = task;
TQ_UNLOCK(queue);
task->ta_func(task->ta_context, pending);
TQ_LOCK(queue);
- *tpp = NULL;
+ tb.tb_running = NULL;
wakeup(task);
}
+ TAILQ_REMOVE(&queue->tq_active, &tb, tb_link);
+}
+
+void
+taskqueue_run(struct taskqueue *queue)
+{
+
+ TQ_LOCK(queue);
+ taskqueue_run_locked(queue);
+ TQ_UNLOCK(queue);
+}
+
+static int
+task_is_running(struct taskqueue *queue, struct task *task)
+{
+ struct taskqueue_busy *tb;
+
+ mtx_assert(&queue->tq_mutex, MA_OWNED);
+ TAILQ_FOREACH(tb, &queue->tq_active, tb_link) {
+ if (tb->tb_running == task)
+ return (1);
+ }
+ return (0);
}
void
@@ -250,10 +283,8 @@ taskqueue_drain(struct taskqueue *queue,
WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL, __func__);
svn commit: r213812 - head/sys/dev/bge
Author: yongari
Date: Wed Oct 13 22:29:48 2010
New Revision: 213812
URL: http://svn.freebsd.org/changeset/base/213812
Log:
Fix a regression introduced in r213710. r213710 removed the use of
auto polling such that it made all controllers obtain link status
information from the state of the LNKRDY input signal. Broadcom
recommends disabling auto polling such that driver should rely on
PHY interrupts for link status change indications. Unfortunately it
seems some controllers(BCM5703, BCM5704 and BCM5705) have PHY
related issues so Linux took other approach to workaround it.
bge(4) didn't follow that and it used to enable auto polling to
workaround it. Restore this old behavior for BCM5700 family
controllers and BCM5705 to use auto polling. For BCM5700 and
BCM5701, it seems it does not need to enable auto polling but I
restored it for safety.
Special thanks to marius who tried lots of patches with patience.
Reported by: marius
Tested by:marius
Modified:
head/sys/dev/bge/if_bge.c
Modified: head/sys/dev/bge/if_bge.c
==
--- head/sys/dev/bge/if_bge.c Wed Oct 13 22:18:03 2010(r213811)
+++ head/sys/dev/bge/if_bge.c Wed Oct 13 22:29:48 2010(r213812)
@@ -2012,6 +2012,10 @@ bge_blockinit(struct bge_softc *sc)
if (sc->bge_flags & BGE_FLAG_TBI) {
CSR_WRITE_4(sc, BGE_MI_STS, BGE_MISTS_LINK);
} else {
+ if (sc->bge_mi_mode & BGE_MIMODE_AUTOPOLL) {
+ CSR_WRITE_4(sc, BGE_MI_MODE, sc->bge_mi_mode);
+ DELAY(80);
+ }
if (sc->bge_asicrev == BGE_ASICREV_BCM5700 &&
sc->bge_chipid != BGE_CHIPID_BCM5700_B2)
CSR_WRITE_4(sc, BGE_MAC_EVT_ENB,
@@ -2677,6 +2681,9 @@ bge_attach(device_t dev)
sc->bge_mi_mode = BGE_MIMODE_500KHZ_CONST;
else
sc->bge_mi_mode = BGE_MIMODE_BASE;
+ /* Enable auto polling for BCM570[0-5]. */
+ if (BGE_IS_5700_FAMILY(sc) || sc->bge_asicrev == BGE_ASICREV_BCM5705)
+ sc->bge_mi_mode |= BGE_MIMODE_AUTOPOLL;
/*
* All controllers that are not 5755 or higher have 4GB
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213744 - head/bin/sh
On Wed, Oct 13, 2010 at 09:07:59AM -0400, John Baldwin wrote: > To be honest, I think changing STATIC is excessive churn. At this point I'm just making everything consistently "static" so that folks that add to ash won't need to realize that "STATIC" exists. Some of the folks making various FreeBSD changes did not realize this and created inconsistencies. Surely this inconsistency isn't desired. > The "right" fix > is to use 'make DEBUG_FLAGS="-g -DDEBUG=2 -fno-inline"'. I often use > 'make DEBUG_FLAGS="-g -fno-inline"' to workaround excessive inlining when I've added "-fno-inline" to DEBUG_FLAGS and will leave it up to others if that spelling isn't accepted by clang, Intel C, or what ever compiler someone may want to use. -- -- David (obr...@freebsd.org) ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213744 - head/bin/sh
On Wed, Oct 13, 2010 at 03:29:27PM +1100, Bruce Evans wrote: > Add __noinline or whatever attributes to STATIC (but keep static in > it) for the DEBUG >= 3 case if you are going that far. __noinline > should be a syntax error for variables, so this should also find any > STATICs still on variables. The spelling fix of changing STATIC to > what it actually means > (ASSORTED_HACKS_FOR_DEBUGGING_BUT_NOW_ONLY_FOR_FUNCTIONS) goes too far > for me. I've make all uses "static". Otherwise we have some of the functions and variable "STATIC" (if from 4.4BSD), and some others "static" (added later by FreeBSD). Surely this inconsistency isn't desired. -- -- David (obr...@freebsd.org) ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213811 - head/bin/sh
Author: obrien
Date: Wed Oct 13 22:18:03 2010
New Revision: 213811
URL: http://svn.freebsd.org/changeset/base/213811
Log:
In the spirit of r90111, depend on c89 and remove the "STATIC" macro
and its usage.
Modified:
head/bin/sh/Makefile
head/bin/sh/alias.c
head/bin/sh/cd.c
head/bin/sh/error.c
head/bin/sh/eval.c
head/bin/sh/exec.c
head/bin/sh/expand.c
head/bin/sh/histedit.c
head/bin/sh/input.c
head/bin/sh/jobs.c
head/bin/sh/main.c
head/bin/sh/memalloc.c
head/bin/sh/nodes.c.pat
head/bin/sh/options.c
head/bin/sh/output.c
head/bin/sh/parser.c
head/bin/sh/redir.c
head/bin/sh/shell.h
head/bin/sh/show.c
head/bin/sh/trap.c
head/bin/sh/var.c
Modified: head/bin/sh/Makefile
==
--- head/bin/sh/MakefileWed Oct 13 22:07:57 2010(r213810)
+++ head/bin/sh/MakefileWed Oct 13 22:18:03 2010(r213811)
@@ -21,7 +21,7 @@ LDADD= -ll -ledit -ltermcap
LFLAGS= -8 # 8-bit lex scanner for arithmetic
CFLAGS+=-DSHELL -I. -I${.CURDIR}
# for debug:
-# DEBUG_FLAGS+= -g -DDEBUG=3 -fno-inline
+# DEBUG_FLAGS+= -g -DDEBUG=2 -fno-inline
WARNS?=2
WFORMAT=0
Modified: head/bin/sh/alias.c
==
--- head/bin/sh/alias.c Wed Oct 13 22:07:57 2010(r213810)
+++ head/bin/sh/alias.c Wed Oct 13 22:18:03 2010(r213811)
@@ -52,11 +52,11 @@ __FBSDID("$FreeBSD$");
static struct alias *atab[ATABSIZE];
static int aliases;
-STATIC void setalias(const char *, const char *);
-STATIC int unalias(const char *);
-STATIC struct alias **hashalias(const char *);
+static void setalias(const char *, const char *);
+static int unalias(const char *);
+static struct alias **hashalias(const char *);
-STATIC
+static
void
setalias(const char *name, const char *val)
{
@@ -111,7 +111,7 @@ setalias(const char *name, const char *v
INTON;
}
-STATIC int
+static int
unalias(const char *name)
{
struct alias *ap, **app;
@@ -191,7 +191,7 @@ lookupalias(const char *name, int check)
return (NULL);
}
-STATIC int
+static int
comparealiases(const void *p1, const void *p2)
{
const struct alias *const *a1 = p1;
@@ -200,7 +200,7 @@ comparealiases(const void *p1, const voi
return strcmp((*a1)->name, (*a2)->name);
}
-STATIC void
+static void
printalias(const struct alias *a)
{
char *p;
@@ -214,7 +214,7 @@ printalias(const struct alias *a)
out1c('\n');
}
-STATIC void
+static void
printaliases(void)
{
int i, j;
@@ -276,7 +276,7 @@ unaliascmd(int argc __unused, char **arg
return (i);
}
-STATIC struct alias **
+static struct alias **
hashalias(const char *p)
{
unsigned int hashval;
Modified: head/bin/sh/cd.c
==
--- head/bin/sh/cd.cWed Oct 13 22:07:57 2010(r213810)
+++ head/bin/sh/cd.cWed Oct 13 22:18:03 2010(r213811)
@@ -64,14 +64,14 @@ __FBSDID("$FreeBSD$");
#include "show.h"
#include "cd.h"
-STATIC int cdlogical(char *);
-STATIC int cdphysical(char *);
-STATIC int docd(char *, int, int);
-STATIC char *getcomponent(void);
-STATIC char *findcwd(char *);
-STATIC void updatepwd(char *);
-STATIC char *getpwd(void);
-STATIC char *getpwd2(void);
+static int cdlogical(char *);
+static int cdphysical(char *);
+static int docd(char *, int, int);
+static char *getcomponent(void);
+static char *findcwd(char *);
+static void updatepwd(char *);
+static char *getpwd(void);
+static char *getpwd2(void);
static char *curdir = NULL;/* current working directory */
static char *prevdir; /* previous working directory */
@@ -145,7 +145,7 @@ cdcmd(int argc, char **argv)
* Actually change the directory. In an interactive shell, print the
* directory name if "print" is nonzero.
*/
-STATIC int
+static int
docd(char *dest, int print, int phys)
{
@@ -161,7 +161,7 @@ docd(char *dest, int print, int phys)
return 0;
}
-STATIC int
+static int
cdlogical(char *dest)
{
char *p;
@@ -213,7 +213,7 @@ cdlogical(char *dest)
return (0);
}
-STATIC int
+static int
cdphysical(char *dest)
{
char *p;
@@ -232,7 +232,7 @@ cdphysical(char *dest)
* Get the next component of the path name pointed to by cdcomppath.
* This routine overwrites the string pointed to by cdcomppath.
*/
-STATIC char *
+static char *
getcomponent(void)
{
char *p;
@@ -253,7 +253,7 @@ getcomponent(void)
}
-STATIC char *
+static char *
findcwd(char *dir)
{
char *new;
@@ -296,7 +296,7 @@ findcwd(char *dir)
* cd command. We also call hashcd to let the routines in exec.c know
* that the current directory has changed.
*/
-STATIC void
+static void
updatepwd(char *dir)
{
hashcd(); /* update command hash table */
@@ -352,7 +352,7 @@ pwdcmd(int argc, char **argv)
/*
svn commit: r213810 - head/sbin/ipfw
Author: luigi Date: Wed Oct 13 22:07:57 2010 New Revision: 213810 URL: http://svn.freebsd.org/changeset/base/213810 Log: document logging through bpf Modified: head/sbin/ipfw/ipfw.8 Modified: head/sbin/ipfw/ipfw.8 == --- head/sbin/ipfw/ipfw.8 Wed Oct 13 22:04:55 2010(r213809) +++ head/sbin/ipfw/ipfw.8 Wed Oct 13 22:07:57 2010(r213810) @@ -557,28 +557,33 @@ packet delivery. Note: this condition is checked before any other condition, including ones such as keep-state or check-state which might have side effects. .It Cm log Op Cm logamount Ar number -When a packet matches a rule with the +Packets matching a rule with the .Cm log -keyword, a message will be -logged to +keyword will be made available for logging in two ways: +if the sysctl variable +.Va net.inet.ip.fw.verbose +is set to 0 (default), one can use +.Xr bpf 4 +attached to the +.Xr ipfw0 +pseudo interface. There is no overhead if no +.Xr bpf +is attached to the pseudo interface. +.Pp +If +.Va net.inet.ip.fw.verbose +is set to 1, packets will be logged to .Xr syslogd 8 with a .Dv LOG_SECURITY -facility. -The logging only occurs if the sysctl variable -.Va net.inet.ip.fw.verbose -is set to 1 -(which is the default when the kernel is compiled with -.Dv IPFIREWALL_VERBOSE ) -and the number of packets logged so far for that -particular rule does not exceed the +facility up to a maximum of .Cm logamount -parameter. +packets. If no .Cm logamount is specified, the limit is taken from the sysctl variable .Va net.inet.ip.fw.verbose_limit . -In both cases, a value of 0 removes the logging limit. +In both cases, a value of 0 means unlimited logging. .Pp Once the limit is reached, logging can be re-enabled by clearing the logging counter or the packet counter for that entry, see the ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213809 - in head/sys/dev/usb: . net
Author: hselasky
Date: Wed Oct 13 22:04:55 2010
New Revision: 213809
URL: http://svn.freebsd.org/changeset/base/213809
Log:
USB network (NCM driver):
- correct the ethernet payload remainder which
must be post-offseted by -14 bytes instead of
0 bytes. This is not very clearly defined in the
NCM specification.
- add development feature about limiting the
maximum datagram count in each NCM payload.
- zero-pad alignment data
- add TX-interval tuning sysctl
Approved by:thompsa (mentor)
Modified:
head/sys/dev/usb/net/if_cdce.c
head/sys/dev/usb/net/if_cdcereg.h
head/sys/dev/usb/usb_cdc.h
Modified: head/sys/dev/usb/net/if_cdce.c
==
--- head/sys/dev/usb/net/if_cdce.c Wed Oct 13 21:53:37 2010
(r213808)
+++ head/sys/dev/usb/net/if_cdce.c Wed Oct 13 22:04:55 2010
(r213809)
@@ -110,10 +110,13 @@ static uint32_t cdce_m_crc32(struct mbuf
#ifdef USB_DEBUG
static int cdce_debug = 0;
+static int cdce_tx_interval = 0;
SYSCTL_NODE(_hw_usb, OID_AUTO, cdce, CTLFLAG_RW, 0, "USB CDC-Ethernet");
SYSCTL_INT(_hw_usb_cdce, OID_AUTO, debug, CTLFLAG_RW, &cdce_debug, 0,
"Debug level");
+SYSCTL_INT(_hw_usb_cdce, OID_AUTO, interval, CTLFLAG_RW, &cdce_tx_interval, 0,
+"NCM transmit interval in ms");
#endif
static const struct usb_config cdce_config[CDCE_N_TRANSFER] = {
@@ -192,7 +195,7 @@ static const struct usb_config cdce_ncm_
.if_index = 0,
.frames = CDCE_NCM_TX_FRAMES_MAX,
.bufsize = (CDCE_NCM_TX_FRAMES_MAX * CDCE_NCM_TX_MAXLEN),
- .flags = {.pipe_bof = 1,.force_short_xfer = 1,},
+ .flags = {.pipe_bof = 1,},
.callback = cdce_ncm_bulk_write_callback,
.timeout = 1, /* 10 seconds */
.usb_mode = USB_MODE_DUAL, /* both modes */
@@ -294,9 +297,22 @@ cdce_ncm_init(struct cdce_softc *sc)
{
struct usb_ncm_parameters temp;
struct usb_device_request req;
- uDWord value;
+ struct usb_ncm_func_descriptor *ufd;
+ uint8_t value[8];
int err;
+ ufd = usbd_find_descriptor(sc->sc_ue.ue_udev, NULL,
+ sc->sc_ifaces_index[1], UDESC_CS_INTERFACE, 0 - 1,
+ UCDC_NCM_FUNC_DESC_SUBTYPE, 0 - 1);
+
+ /* verify length of NCM functional descriptor */
+ if (ufd != NULL) {
+ if (ufd->bLength < sizeof(*ufd))
+ ufd = NULL;
+ else
+ DPRINTFN(1, "Found NCM functional descriptor.\n");
+ }
+
req.bmRequestType = UT_READ_CLASS_INTERFACE;
req.bRequest = UCDC_NCM_GET_NTB_PARAMETERS;
USETW(req.wValue, 0);
@@ -317,17 +333,19 @@ cdce_ncm_init(struct cdce_softc *sc)
sc->sc_ncm.tx_remainder = UGETW(temp.wNdpOutPayloadRemainder);
sc->sc_ncm.tx_modulus = UGETW(temp.wNdpOutDivisor);
sc->sc_ncm.tx_struct_align = UGETW(temp.wNdpOutAlignment);
+ sc->sc_ncm.tx_nframe = UGETW(temp.wNtbOutMaxDatagrams);
} else {
sc->sc_ncm.rx_max = UGETDW(temp.dwNtbOutMaxSize);
sc->sc_ncm.tx_max = UGETDW(temp.dwNtbInMaxSize);
sc->sc_ncm.tx_remainder = UGETW(temp.wNdpInPayloadRemainder);
sc->sc_ncm.tx_modulus = UGETW(temp.wNdpInDivisor);
sc->sc_ncm.tx_struct_align = UGETW(temp.wNdpInAlignment);
+ sc->sc_ncm.tx_nframe = UGETW(temp.wNtbOutMaxDatagrams);
}
/* Verify maximum receive length */
- if (err || (sc->sc_ncm.rx_max < 32) ||
+ if ((sc->sc_ncm.rx_max < 32) ||
(sc->sc_ncm.rx_max > CDCE_NCM_RX_MAXLEN)) {
DPRINTFN(1, "Using default maximum receive length\n");
sc->sc_ncm.rx_max = CDCE_NCM_RX_MAXLEN;
@@ -335,7 +353,7 @@ cdce_ncm_init(struct cdce_softc *sc)
/* Verify maximum transmit length */
- if (err || (sc->sc_ncm.tx_max < 32) ||
+ if ((sc->sc_ncm.tx_max < 32) ||
(sc->sc_ncm.tx_max > CDCE_NCM_TX_MAXLEN)) {
DPRINTFN(1, "Using default maximum transmit length\n");
sc->sc_ncm.tx_max = CDCE_NCM_TX_MAXLEN;
@@ -347,7 +365,7 @@ cdce_ncm_init(struct cdce_softc *sc)
* - not greater than the maximum transmit length
* - not less than four bytes
*/
- if (err || (sc->sc_ncm.tx_struct_align < 4) ||
+ if ((sc->sc_ncm.tx_struct_align < 4) ||
(sc->sc_ncm.tx_struct_align !=
((-sc->sc_ncm.tx_struct_align) & sc->sc_ncm.tx_struct_align)) ||
(sc->sc_ncm.tx_struct_align >= sc->sc_ncm.tx_max)) {
@@ -361,7 +379,7 @@ cdce_ncm_init(struct cdce_softc *sc)
* - not greater than the maximum transmit length
* - not less than four bytes
*/
- if (err || (sc->sc_ncm.tx_modulus < 4) ||
+ if ((sc->sc_ncm.tx_modulus < 4) ||
(sc->sc_ncm.tx_m
svn commit: r213808 - head/sys/dev/bge
Author: yongari
Date: Wed Oct 13 21:53:37 2010
New Revision: 213808
URL: http://svn.freebsd.org/changeset/base/213808
Log:
Add more checks for resolved link speed in bge_miibus_statchg().
Link UP state could be reported first before actual completion of
auto-negotiation. This change makes bge(4) reprogram BGE_MAC_MODE,
BGE_TX_MODE and BGE_RX_MODE register only after controller got a
valid link.
Modified:
head/sys/dev/bge/if_bge.c
Modified: head/sys/dev/bge/if_bge.c
==
--- head/sys/dev/bge/if_bge.c Wed Oct 13 21:45:56 2010(r213807)
+++ head/sys/dev/bge/if_bge.c Wed Oct 13 21:53:37 2010(r213808)
@@ -878,6 +878,29 @@ bge_miibus_statchg(device_t dev)
sc = device_get_softc(dev);
mii = device_get_softc(sc->bge_miibus);
+ if ((mii->mii_media_status & (IFM_ACTIVE | IFM_AVALID)) ==
+ (IFM_ACTIVE | IFM_AVALID)) {
+ switch (IFM_SUBTYPE(mii->mii_media_active)) {
+ case IFM_10_T:
+ case IFM_100_TX:
+ sc->bge_link = 1;
+ break;
+ case IFM_1000_T:
+ case IFM_1000_SX:
+ case IFM_2500_SX:
+ if (sc->bge_asicrev != BGE_ASICREV_BCM5906)
+ sc->bge_link = 1;
+ else
+ sc->bge_link = 0;
+ break;
+ default:
+ sc->bge_link = 0;
+ break;
+ }
+ } else
+ sc->bge_link = 0;
+ if (sc->bge_link == 0)
+ return;
BGE_CLRBIT(sc, BGE_MAC_MODE, BGE_MACMODE_PORTMODE);
if (IFM_SUBTYPE(mii->mii_media_active) == IFM_1000_T ||
IFM_SUBTYPE(mii->mii_media_active) == IFM_1000_SX)
@@ -5065,12 +5088,7 @@ bge_link_upd(struct bge_softc *sc)
*/
mii = device_get_softc(sc->bge_miibus);
mii_pollstat(mii);
- if (!sc->bge_link && mii->mii_media_status & IFM_ACTIVE &&
- IFM_SUBTYPE(mii->mii_media_active) != IFM_NONE) {
- bge_miibus_statchg(sc->bge_dev);
- sc->bge_link = 1;
- } else
- sc->bge_link = 0;
+ bge_miibus_statchg(sc->bge_dev);
}
/* Clear the attention. */
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213807 - head/sys/mips/cavium/octe
Author: jmallett
Date: Wed Oct 13 21:45:56 2010
New Revision: 213807
URL: http://svn.freebsd.org/changeset/base/213807
Log:
Keep polling at 50hz as long as link state is changing.
Modified:
head/sys/mips/cavium/octe/ethernet.c
Modified: head/sys/mips/cavium/octe/ethernet.c
==
--- head/sys/mips/cavium/octe/ethernet.cWed Oct 13 21:37:02 2010
(r213806)
+++ head/sys/mips/cavium/octe/ethernet.cWed Oct 13 21:45:56 2010
(r213807)
@@ -173,6 +173,7 @@ static void cvm_oct_update_link(void *co
static void cvm_do_timer(void *arg)
{
static int port;
+ static int updated;
if (port < CVMX_PIP_NUM_INPUT_PORTS) {
if (cvm_oct_device[port]) {
int queues_per_port;
@@ -186,6 +187,7 @@ static void cvm_do_timer(void *arg)
MDIO_UNLOCK();
if (priv->need_link_update) {
+ updated++;
taskqueue_enqueue(cvm_oct_link_taskq, &priv->link_task);
}
}
@@ -220,9 +222,19 @@ static void cvm_do_timer(void *arg)
callout_reset(&cvm_oct_poll_timer, hz / 50, cvm_do_timer, NULL);
} else {
port = 0;
- /* All ports have been polled. Start the next iteration through
- the ports in one second */
- callout_reset(&cvm_oct_poll_timer, hz, cvm_do_timer, NULL);
+ /* If any updates were made in this run, continue iterating at
+* 1/50th of a second, so that if a link has merely gone down
+* temporarily (e.g. because of interface reinitialization) it
+* will not be forced to stay down for an entire second.
+*/
+ if (updated > 0) {
+ updated = 0;
+ callout_reset(&cvm_oct_poll_timer, hz / 50,
cvm_do_timer, NULL);
+ } else {
+ /* All ports have been polled. Start the next iteration
through
+ the ports in one second */
+ callout_reset(&cvm_oct_poll_timer, hz, cvm_do_timer,
NULL);
+ }
}
}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213779 - head/sys/dev/sound/pci
On Wednesday 13 October 2010 05:17 pm, Rui Paulo wrote: > On 13 Oct 2010, at 22:03, Jung-uk Kim wrote: > > On Wednesday 13 October 2010 04:30 pm, you wrote: > >> I read the PR and the mailing list posts, but I don't see what > >> problem does "case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:" > >> fix. > > > > A similar bug for right channel was fixed by netchild@ in > > r188480: > > > > http://svn.freebsd.org/viewvc/base/head/sys/dev/sound/pci/spicds. > >c?view=log > > > > but he missed the left channel, which caused volume differences > > between the two channels. > > > > "In FreeBSD 7.2 it worked just like a charm but after upgrading > > to 8.0 the left stereo channel is only half as loud as the right > > one. It can be reproduced with either speakers and headphones. I > > tracked it down to a change in revision 188480 of spicds.c, the > > change "fix: stupid bug with volume control for AK4396" breaks > > volume control for me. > > > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-Decemb > >er/010553html > > > > Carl Johan Gustavsson submitted a correct patch: > > > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-Decemb > >er/010558html > > > > ariff@ said he would take it but I guess he never did: > > > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-Decemb > >er/010575html > > Oh, I read your previous email backwards. This explains it, thanks. :-) > Guess we can close the PR now. Yes, that's exactly what I wanted to tell you. Thanks! Jung-uk Kim ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213806 - in head/sys/contrib/dev/acpica: . common compiler events include include/platform utilities
Author: jkim Date: Wed Oct 13 21:37:02 2010 New Revision: 213806 URL: http://svn.freebsd.org/changeset/base/213806 Log: Merge ACPICA 20101013. Added: head/sys/contrib/dev/acpica/compiler/aslmessages.h - copied unchanged from r213804, vendor-sys/acpica/dist/compiler/aslmessages.h Modified: head/sys/contrib/dev/acpica/acpica_prep.sh head/sys/contrib/dev/acpica/changes.txt head/sys/contrib/dev/acpica/common/adisasm.c head/sys/contrib/dev/acpica/compiler/aslanalyze.c head/sys/contrib/dev/acpica/compiler/aslcodegen.c head/sys/contrib/dev/acpica/compiler/aslcompile.c head/sys/contrib/dev/acpica/compiler/aslcompiler.h head/sys/contrib/dev/acpica/compiler/asldefine.h head/sys/contrib/dev/acpica/compiler/aslmain.c head/sys/contrib/dev/acpica/compiler/aslresource.c head/sys/contrib/dev/acpica/compiler/aslrestype1.c head/sys/contrib/dev/acpica/compiler/aslrestype1i.c head/sys/contrib/dev/acpica/compiler/aslrestype2d.c head/sys/contrib/dev/acpica/compiler/aslrestype2e.c head/sys/contrib/dev/acpica/compiler/aslrestype2q.c head/sys/contrib/dev/acpica/compiler/aslrestype2w.c head/sys/contrib/dev/acpica/compiler/asltypes.h head/sys/contrib/dev/acpica/compiler/aslutils.c head/sys/contrib/dev/acpica/compiler/dtcompile.c head/sys/contrib/dev/acpica/compiler/dttemplate.c head/sys/contrib/dev/acpica/events/evxfregn.c head/sys/contrib/dev/acpica/include/acapps.h head/sys/contrib/dev/acpica/include/aclocal.h head/sys/contrib/dev/acpica/include/acpixf.h head/sys/contrib/dev/acpica/include/platform/acenv.h head/sys/contrib/dev/acpica/osunixxf.c head/sys/contrib/dev/acpica/utilities/utglobal.c head/sys/contrib/dev/acpica/utilities/utids.c head/sys/contrib/dev/acpica/utilities/utosi.c Directory Properties: head/sys/contrib/dev/acpica/ (props changed) Modified: head/sys/contrib/dev/acpica/acpica_prep.sh == --- head/sys/contrib/dev/acpica/acpica_prep.sh Wed Oct 13 21:36:42 2010 (r213805) +++ head/sys/contrib/dev/acpica/acpica_prep.sh Wed Oct 13 21:37:02 2010 (r213806) @@ -22,8 +22,8 @@ fulldirs="common compiler debugger disas stripdirs="acpisrc acpixtract examples generate os_specific tests" stripfiles="Makefile README acintel.h aclinux.h acmsvc.h acnetbsd.h\ acos2.h accygwin.h acefi.h acwin.h acwin64.h aeexec.c \ - aehandlers.c aemain.c aetables.c osunixdir.c readme.txt \ - utclib.c" + aehandlers.c aemain.c aetables.c aetables.h osunixdir.c \ + readme.txt utclib.c" # include files to canonify src_headers="acapps.h accommon.h acconfig.h acdebug.h acdisasm.h \ @@ -33,8 +33,8 @@ src_headers="acapps.h accommon.h acconfi acresrc.h acrestyp.h acstruct.h actables.h actbl.h actbl1.h \ actbl2.h actypes.h acutils.h amlcode.h amlresrc.h \ platform/acenv.h platform/acfreebsd.h platform/acgcc.h" -comp_headers="aslcompiler.h asldefine.h aslglobal.h asltypes.h \ - dtcompiler.h dttemplate.h" +comp_headers="aslcompiler.h asldefine.h aslglobal.h aslmessages.h \ + asltypes.h dtcompiler.h dttemplate.h" platform_headers="acfreebsd.h acgcc.h" # pre-clean Modified: head/sys/contrib/dev/acpica/changes.txt == --- head/sys/contrib/dev/acpica/changes.txt Wed Oct 13 21:36:42 2010 (r213805) +++ head/sys/contrib/dev/acpica/changes.txt Wed Oct 13 21:37:02 2010 (r213806) @@ -1,4 +1,64 @@ ---- +13 October 2010. Summary of changes for version 20101013: + +This release is available at www.acpica.org/downloads + +1) ACPI CA Core Subsystem: + +Added support to clear the PCIEXP_WAKE event. When clearing ACPI events, now +clear the PCIEXP_WAKE_STS bit in the ACPI PM1 Status Register, via +HwClearAcpiStatus. Original change from Colin King. ACPICA BZ 880. + +Changed the type of the predefined namespace object _TZ from ThermalZone to +Device. This was found to be confusing to the host software that processes +the various thermal zones, since _TZ is not really a ThermalZone. However, a +Notify() can still be performed on it. ACPICA BZ 876. Suggestion from Rui +Zhang. + +Added Windows Vista SP2 to the list of supported _OSI strings. The actual +string is "Windows 2006 SP2". + +Eliminated duplicate code in AcpiUtExecute* functions. Now that the nsrepair +code automatically repairs _HID-related strings, this type of code is no +longer needed in Execute_HID, Execute_CID, and Execute_UID. ACPICA BZ 878. + +Example Code and Data Size: These are the sizes for the OS-independent +acpica.lib produced by the Microsoft Visual C++ 6.0 32-bit compiler. The +debug version of the code includes the debug output trace mechanism and has a +much larger
svn commit: r213805 - in head/sys: conf dev/usb/net modules/usb modules/usb/ipheth
Author: hselasky
Date: Wed Oct 13 21:36:42 2010
New Revision: 213805
URL: http://svn.freebsd.org/changeset/base/213805
Log:
USB Network:
- Add new driver for iPhone tethering
- Supports the iPhone 3G/3GS/4G ethernet protocol
Approved by:thompsa (mentor)
Added:
head/sys/dev/usb/net/if_ipheth.c (contents, props changed)
head/sys/dev/usb/net/if_iphethvar.h (contents, props changed)
head/sys/modules/usb/ipheth/
head/sys/modules/usb/ipheth/Makefile (contents, props changed)
Modified:
head/sys/conf/files
head/sys/modules/usb/Makefile
Modified: head/sys/conf/files
==
--- head/sys/conf/files Wed Oct 13 20:56:54 2010(r213804)
+++ head/sys/conf/files Wed Oct 13 21:36:42 2010(r213805)
@@ -1790,6 +1790,7 @@ dev/usb/net/if_aue.c optional aue
dev/usb/net/if_axe.c optional axe
dev/usb/net/if_cdce.c optional cdce
dev/usb/net/if_cue.c optional cue
+dev/usb/net/if_ipheth.coptional ipheth
dev/usb/net/if_kue.c optional kue
dev/usb/net/if_rue.c optional rue
dev/usb/net/if_udav.c optional udav
Added: head/sys/dev/usb/net/if_ipheth.c
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/sys/dev/usb/net/if_ipheth.cWed Oct 13 21:36:42 2010
(r213805)
@@ -0,0 +1,528 @@
+/*-
+ * Copyright (c) 2010 Hans Petter Selasky. All rights reserved.
+ * Copyright (c) 2009 Diego Giagio. 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 AND CONTRIBUTORS ``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 OR CONTRIBUTORS 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.
+ */
+
+/*
+ * Thanks to Diego Giagio for figuring out the programming details for
+ * the Apple iPhone Ethernet driver.
+ */
+
+#include
+__FBSDID("$FreeBSD$");
+
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+
+#include
+#include
+#include
+#include "usbdevs.h"
+
+#defineUSB_DEBUG_VAR ipheth_debug
+#include
+#include
+
+#include
+#include
+
+static device_probe_t ipheth_probe;
+static device_attach_t ipheth_attach;
+static device_detach_t ipheth_detach;
+
+static usb_callback_t ipheth_bulk_write_callback;
+static usb_callback_t ipheth_bulk_read_callback;
+
+static uether_fn_t ipheth_attach_post;
+static uether_fn_t ipheth_tick;
+static uether_fn_t ipheth_init;
+static uether_fn_t ipheth_stop;
+static uether_fn_t ipheth_start;
+static uether_fn_t ipheth_setmulti;
+static uether_fn_t ipheth_setpromisc;
+
+#ifdef USB_DEBUG
+static int ipheth_debug = 0;
+
+SYSCTL_NODE(_hw_usb, OID_AUTO, ipheth, CTLFLAG_RW, 0, "USB iPhone ethernet");
+SYSCTL_INT(_hw_usb_ipheth, OID_AUTO, debug, CTLFLAG_RW, &ipheth_debug, 0,
"Debug level");
+#endif
+
+static const struct usb_config ipheth_config[IPHETH_N_TRANSFER] = {
+
+ [IPHETH_BULK_RX] = {
+ .type = UE_BULK,
+ .endpoint = UE_ADDR_ANY,
+ .direction = UE_DIR_RX,
+ .frames = IPHETH_RX_FRAMES_MAX,
+ .bufsize = (IPHETH_RX_FRAMES_MAX * MCLBYTES),
+ .flags = {.short_frames_ok = 1,.short_xfer_ok = 1,.ext_buffer =
1,},
+ .callback = ipheth_bulk_read_callback,
+ .timeout = 0, /* no timeout */
+ },
+
+ [IPHETH_BULK_TX] = {
+ .type = UE_BULK,
+ .endpoint = UE_ADDR_ANY,
+ .direction = UE_DIR_TX,
+ .frames = IPHETH_TX_FRAMES_MAX,
+ .bufsize = (IPHETH_TX_FRAMES_MAX * IPHETH_BUF_SIZE),
+ .flags = {.force_short_xfer
Re: svn commit: r213779 - head/sys/dev/sound/pci
On 13 Oct 2010, at 22:03, Jung-uk Kim wrote: > On Wednesday 13 October 2010 04:30 pm, you wrote: >> I read the PR and the mailing list posts, but I don't see what >> problem does "case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:" fix. > > A similar bug for right channel was fixed by netchild@ in r188480: > > http://svn.freebsd.org/viewvc/base/head/sys/dev/sound/pci/spicds.c?view=log > > but he missed the left channel, which caused volume differences > between the two channels. > > "In FreeBSD 7.2 it worked just like a charm but after upgrading to 8.0 > the left stereo channel is only half as loud as the right one. It can > be reproduced with either speakers and headphones. I tracked it down > to a change in revision 188480 of spicds.c, the change "fix: stupid > bug with volume control for AK4396" breaks volume control for me. > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010553html > > Carl Johan Gustavsson submitted a correct patch: > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010558html > > ariff@ said he would take it but I guess he never did: > > http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010575html Oh, I read your previous email backwards. This explains it, thanks. Guess we can close the PR now. Regards, -- Rui Paulo ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213779 - head/sys/dev/sound/pci
On Wednesday 13 October 2010 04:30 pm, you wrote: > I read the PR and the mailing list posts, but I don't see what > problem does "case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:" fix. A similar bug for right channel was fixed by netchild@ in r188480: http://svn.freebsd.org/viewvc/base/head/sys/dev/sound/pci/spicds.c?view=log but he missed the left channel, which caused volume differences between the two channels. "In FreeBSD 7.2 it worked just like a charm but after upgrading to 8.0 the left stereo channel is only half as loud as the right one. It can be reproduced with either speakers and headphones. I tracked it down to a change in revision 188480 of spicds.c, the change "fix: stupid bug with volume control for AK4396" breaks volume control for me. http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010553.html Carl Johan Gustavsson submitted a correct patch: http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010558.html ariff@ said he would take it but I guess he never did: http://lists.freebsd.org/pipermail/freebsd-multimedia/2009-December/010575.html Jung-uk Kim ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213804 - head/sys/dev/usb/wlan
Author: hselasky Date: Wed Oct 13 20:56:54 2010 New Revision: 213804 URL: http://svn.freebsd.org/changeset/base/213804 Log: USB WLAN: - Add new device ID PR: usb/150989 Approved by:thompsa (mentor) Modified: head/sys/dev/usb/wlan/if_upgt.c Modified: head/sys/dev/usb/wlan/if_upgt.c == --- head/sys/dev/usb/wlan/if_upgt.c Wed Oct 13 20:51:06 2010 (r213803) +++ head/sys/dev/usb/wlan/if_upgt.c Wed Oct 13 20:56:54 2010 (r213804) @@ -182,6 +182,7 @@ static const struct usb_device_id upgt_d UPGT_DEV(FSC, E5400), UPGT_DEV(GLOBESPAN, PRISM_GT_1), UPGT_DEV(GLOBESPAN, PRISM_GT_2), + UPGT_DEV(NETGEAR, WG111V2_2), UPGT_DEV(INTERSIL, PRISM_GT), UPGT_DEV(SMC, 2862WG), UPGT_DEV(USR, USR5422), ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213803 - head/sys/dev/usb/net
Author: hselasky Date: Wed Oct 13 20:51:06 2010 New Revision: 213803 URL: http://svn.freebsd.org/changeset/base/213803 Log: USB network (UHSO): - Correct network interface flags. PR: usb/149039 Submitted by: Fredrik Lindberg Approved by:thompsa (mentor) Modified: head/sys/dev/usb/net/uhso.c Modified: head/sys/dev/usb/net/uhso.c == --- head/sys/dev/usb/net/uhso.c Wed Oct 13 20:37:19 2010(r213802) +++ head/sys/dev/usb/net/uhso.c Wed Oct 13 20:51:06 2010(r213803) @@ -1560,7 +1560,7 @@ uhso_attach_ifnet(struct uhso_softc *sc, ifp->if_init = uhso_if_init; ifp->if_start = uhso_if_start; ifp->if_output = uhso_if_output; - ifp->if_flags = 0; + ifp->if_flags = IFF_BROADCAST | IFF_MULTICAST | IFF_NOARP; ifp->if_softc = sc; IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen); ifp->if_snd.ifq_drv_maxlen = ifqmaxlen; ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213802 - head/sys/dev/usb/controller
Author: hselasky
Date: Wed Oct 13 20:37:19 2010
New Revision: 213802
URL: http://svn.freebsd.org/changeset/base/213802
Log:
Correct some root HUB descriptor fields in multiple controller drivers.
Remove an unused structure.
Approved by:thompsa (mentor)
Modified:
head/sys/dev/usb/controller/at91dci.c
head/sys/dev/usb/controller/atmegadci.c
head/sys/dev/usb/controller/avr32dci.c
head/sys/dev/usb/controller/musb_otg.c
head/sys/dev/usb/controller/ohci.c
head/sys/dev/usb/controller/uss820dci.c
Modified: head/sys/dev/usb/controller/at91dci.c
==
--- head/sys/dev/usb/controller/at91dci.c Wed Oct 13 20:36:42 2010
(r213801)
+++ head/sys/dev/usb/controller/at91dci.c Wed Oct 13 20:37:19 2010
(r213802)
@@ -1689,7 +1689,7 @@ static const struct usb_device_descripto
.bcdUSB = {0x00, 0x02},
.bDeviceClass = UDCLASS_HUB,
.bDeviceSubClass = UDSUBCLASS_HUB,
- .bDeviceProtocol = UDPROTO_HSHUBSTT,
+ .bDeviceProtocol = UDPROTO_FSHUB,
.bMaxPacketSize = 64,
.bcdDevice = {0x00, 0x01},
.iManufacturer = 1,
@@ -1697,17 +1697,6 @@ static const struct usb_device_descripto
.bNumConfigurations = 1,
};
-static const struct usb_device_qualifier at91dci_odevd = {
- .bLength = sizeof(struct usb_device_qualifier),
- .bDescriptorType = UDESC_DEVICE_QUALIFIER,
- .bcdUSB = {0x00, 0x02},
- .bDeviceClass = UDCLASS_HUB,
- .bDeviceSubClass = UDSUBCLASS_HUB,
- .bDeviceProtocol = UDPROTO_FSHUB,
- .bMaxPacketSize0 = 0,
- .bNumConfigurations = 0,
-};
-
static const struct at91dci_config_desc at91dci_confd = {
.confd = {
.bLength = sizeof(struct usb_config_descriptor),
@@ -1725,7 +1714,7 @@ static const struct at91dci_config_desc
.bNumEndpoints = 1,
.bInterfaceClass = UICLASS_HUB,
.bInterfaceSubClass = UISUBCLASS_HUB,
- .bInterfaceProtocol = UIPROTO_HSHUBSTT,
+ .bInterfaceProtocol = 0,
},
.endpd = {
.bLength = sizeof(struct usb_endpoint_descriptor),
Modified: head/sys/dev/usb/controller/atmegadci.c
==
--- head/sys/dev/usb/controller/atmegadci.c Wed Oct 13 20:36:42 2010
(r213801)
+++ head/sys/dev/usb/controller/atmegadci.c Wed Oct 13 20:37:19 2010
(r213802)
@@ -1511,7 +1511,7 @@ static const struct usb_device_descripto
.bcdUSB = {0x00, 0x02},
.bDeviceClass = UDCLASS_HUB,
.bDeviceSubClass = UDSUBCLASS_HUB,
- .bDeviceProtocol = UDPROTO_HSHUBSTT,
+ .bDeviceProtocol = UDPROTO_FSHUB,
.bMaxPacketSize = 64,
.bcdDevice = {0x00, 0x01},
.iManufacturer = 1,
@@ -1519,17 +1519,6 @@ static const struct usb_device_descripto
.bNumConfigurations = 1,
};
-static const struct usb_device_qualifier atmegadci_odevd = {
- .bLength = sizeof(struct usb_device_qualifier),
- .bDescriptorType = UDESC_DEVICE_QUALIFIER,
- .bcdUSB = {0x00, 0x02},
- .bDeviceClass = UDCLASS_HUB,
- .bDeviceSubClass = UDSUBCLASS_HUB,
- .bDeviceProtocol = UDPROTO_FSHUB,
- .bMaxPacketSize0 = 0,
- .bNumConfigurations = 0,
-};
-
static const struct atmegadci_config_desc atmegadci_confd = {
.confd = {
.bLength = sizeof(struct usb_config_descriptor),
@@ -1547,7 +1536,7 @@ static const struct atmegadci_config_des
.bNumEndpoints = 1,
.bInterfaceClass = UICLASS_HUB,
.bInterfaceSubClass = UISUBCLASS_HUB,
- .bInterfaceProtocol = UIPROTO_HSHUBSTT,
+ .bInterfaceProtocol = 0,
},
.endpd = {
.bLength = sizeof(struct usb_endpoint_descriptor),
Modified: head/sys/dev/usb/controller/avr32dci.c
==
--- head/sys/dev/usb/controller/avr32dci.c Wed Oct 13 20:36:42 2010
(r213801)
+++ head/sys/dev/usb/controller/avr32dci.c Wed Oct 13 20:37:19 2010
(r213802)
@@ -1480,7 +1480,7 @@ static const struct avr32dci_config_desc
.bNumEndpoints = 1,
.bInterfaceClass = UICLASS_HUB,
.bInterfaceSubClass = UISUBCLASS_HUB,
- .bInterfaceProtocol = UIPROTO_HSHUBSTT,
+ .bInterfaceProtocol = 0,
},
.endpd = {
.bLength = sizeof(struct usb_endpoint_descriptor),
Modified: head/sys/dev/usb/controller/musb_otg.c
==
--- head/sys/dev/usb/controller/musb_otg.c Wed Oct 13 20:36:42 2010
(r213801)
+++ head/sys/dev/usb/controller/musb_otg.c Wed Oct 13 20:37:19 2010
(r213802)
@@ -2181,7 +2181,7 @@ static const struct musbotg_config_desc
Re: svn commit: r213779 - head/sys/dev/sound/pci
On 13 Oct 2010, at 20:46, Jung-uk Kim wrote:
> On Wednesday 13 October 2010 10:39 am, Rui Paulo wrote:
>> Author: rpaulo
>> Date: Wed Oct 13 14:39:54 2010
>> New Revision: 213779
>> URL: http://svn.freebsd.org/changeset/base/213779
>>
>> Log:
>> Fix a brain-o: wrong case statement semantics.
>>
>> Found with: clang
>>
>> Modified:
>> head/sys/dev/sound/pci/envy24ht.c
>> head/sys/dev/sound/pci/spicds.c
>>
>> Modified: head/sys/dev/sound/pci/envy24ht.c
>> ===
>> === --- head/sys/dev/sound/pci/envy24ht.cWed Oct 13
>> 14:37:52 2010(r213778) +++ head/sys/dev/sound/pci/envy24ht.c Wed
>> Oct 13 14:39:54 2010 (r213779) @@ -2236,7 +2236,8 @@
>> envy24ht_putcfg(struct sc_info *sc)
>> else
>> printf("not implemented\n");
>> switch (sc->adcn) {
>> -case 0x01 || 0x02:
>> +case 0x01:
>> +case 0x02:
>> printf(" ADC #: ");
>> printf("%d\n", sc->adcn);
>> break;
>>
>> Modified: head/sys/dev/sound/pci/spicds.c
>> ===
>> === --- head/sys/dev/sound/pci/spicds.c Wed Oct 13 14:37:52
>> 2010 (r213778) +++ head/sys/dev/sound/pci/spicds.c Wed Oct 13
>> 14:39:54 2010(r213779) @@ -283,7 +283,8 @@ spicds_set(struct
>> spicds_info *codec, in case SPICDS_TYPE_WM8770:
>> left = left + 27;
>> break;
>> -case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:
>> +case SPICDS_TYPE_AK4381:
>> +case SPICDS_TYPE_AK4396:
>> left = left * 255 / 100;
>> break;
>> default:
>
> Although it was rediscovered by clang, spicds.c patch actually known
> to fix a problem. Please see kern/146758.
I read the PR and the mailing list posts, but I don't see what problem does
"case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:" fix.
Regards,
--
Rui Paulo
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213799 - head/usr.bin/truss
Author: bcr (doc committer) Date: Wed Oct 13 20:08:02 2010 New Revision: 213799 URL: http://svn.freebsd.org/changeset/base/213799 Log: s/sytem/system in comments, no functional changes. Reviewed by:alfred@ Modified: head/usr.bin/truss/amd64-fbsd.c head/usr.bin/truss/amd64-fbsd32.c head/usr.bin/truss/i386-fbsd.c head/usr.bin/truss/ia64-fbsd.c head/usr.bin/truss/mips-fbsd.c head/usr.bin/truss/powerpc-fbsd.c head/usr.bin/truss/powerpc64-fbsd.c head/usr.bin/truss/sparc64-fbsd.c Modified: head/usr.bin/truss/amd64-fbsd.c == --- head/usr.bin/truss/amd64-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/amd64-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -257,7 +257,7 @@ amd64_syscall_entry(struct trussinfo *tr * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/amd64-fbsd32.c == --- head/usr.bin/truss/amd64-fbsd32.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/amd64-fbsd32.c Wed Oct 13 20:08:02 2010 (r213799) @@ -260,7 +260,7 @@ amd64_fbsd32_syscall_entry(struct trussi * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/i386-fbsd.c == --- head/usr.bin/truss/i386-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/i386-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -250,7 +250,7 @@ i386_syscall_entry(struct trussinfo *tru * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/ia64-fbsd.c == --- head/usr.bin/truss/ia64-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/ia64-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -231,7 +231,7 @@ ia64_syscall_entry(struct trussinfo *tru * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/mips-fbsd.c == --- head/usr.bin/truss/mips-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/mips-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -276,7 +276,7 @@ mips_syscall_entry(struct trussinfo *tru * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/powerpc-fbsd.c == --- head/usr.bin/truss/powerpc-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/powerpc-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -262,7 +262,7 @@ powerpc_syscall_entry(struct trussinfo * * And when the system call is done, we handle it here. * Currently, no attempt is made to ensure that the system calls * match -- this needs to be fixed (and is, in fact, why S_SCX includes - * the sytem call number instead of, say, an error status). + * the system call number instead of, say, an error status). */ long Modified: head/usr.bin/truss/powerpc64-fbsd.c == --- head/usr.bin/truss/powerpc64-fbsd.c Wed Oct 13 18:23:43 2010 (r213798) +++ head/usr.bin/truss/powerpc64-fbsd.c Wed Oct 13 20:08:02 2010 (r213799) @@ -250,7 +250,7 @@ powerpc64_syscall_entry(struct trussinfo * And when the system call is done, we handle it
Re: svn commit: r213793 - in head/sys/dev: ce cp
On Wednesday, October 13, 2010 1:59:42 pm Jung-uk Kim wrote:
> On Wednesday 13 October 2010 01:27 pm, Roman Divacky wrote:
> > On Wed, Oct 13, 2010 at 05:17:50PM +, Rui Paulo wrote:
> > > Author: rpaulo
> > > Date: Wed Oct 13 17:17:50 2010
> > > New Revision: 213793
> > > URL: http://svn.freebsd.org/changeset/base/213793
> > >
> > > Log:
> > > Don't do a logical AND of the result of strcmp() with a
> > > constant.
> > >
> > > Found with: clang
> > >
> > > Modified:
> > > head/sys/dev/ce/if_ce.c
> > > head/sys/dev/cp/if_cp.c
> > >
> > > Modified: head/sys/dev/ce/if_ce.c
> > > =
> > >= --- head/sys/dev/ce/if_ce.c Wed Oct 13 17:16:08
> > > 2010 (r213792) +++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:17:50
> > > 2010 (r213793) @@ -1313,7 +1313,7 @@ static int ce_ioctl (struct
> > > cdev *dev, u IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
> > > IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
> > > d->ifp->if_flags |= PP_CISCO;
> > > - } else if (! strcmp ("fr", (char*)data) && PP_FR) {
> > > + } else if (! strcmp ("fr", (char*)data)) {
> >
> > this is wrong I think... the PP_FR was used for compiling in/out
> > support for something.. see the comment:
> >
> > /* If we don't have Cronyx's sppp version, we don't have fr support
> > via sppp */ #ifndef PP_FR
> > #define PP_FR 0
> > #endif
> >
> > note that PP_FR is used in some other places as a flag. I guess
> > that by compiling with something like make -DPP_FR=42 some magic
> > would happen.
> >
> > anyway - this does not look like a bug but like an intent, please
> > revert.
>
> I think the attached patch should do.
This is certainly more obvious and readable.
--
John Baldwin
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213648 - head/sys/kern
On Wednesday, October 13, 2010 12:21:44 pm Gennady Proskurin wrote:
> Thank you and Bruce for explanation.
> The key point here (which I did not understand) is that
> "something == PCPU_GET(cpuid))" may be true only if "something" is set by the
> current cpu, in which case its value is not stale.
Yes. mtx_owned() (and mtx_assert()) depend on a similar "feature" in that
stale values don't hurt.
> On Wed, Oct 13, 2010 at 09:05:04AM -0400, John Baldwin wrote:
> > On Tuesday, October 12, 2010 5:08:04 pm Gennady Proskurin wrote:
> > > On Sat, Oct 09, 2010 at 12:48:50PM +0300, Andriy Gapon wrote:
> > > > on 09/10/2010 12:33 Bruce Evans said the following:
> > > > > On Sat, 9 Oct 2010, Andriy Gapon wrote:
> > > > >
> > > > >> Log:
> > > > >> panic_cpu variable should be volatile
> > > > >>
> > > > >> This is to prevent caching of its value in a register when it is
> > > > >> checked
> > > > >> and modified by multiple CPUs in parallel.
> > > > >> Also, move the variable into the scope of the only function that
> > > > >> uses it.
> > > > >>
> > > > >> Reviewed by:jhb
> > > > >> Hint from:mdf
> > > > >> MFC after:1 week
> > > > >
> > > > > I doubt that this is either necessary or sufficient. Most variables
> > > > > aren't volatile while they are locked by a mutex. But panic() cannot
> > > > > use
> > > > > normal mutex locking, so it should access this variable using atomic
> > > > > ops with memory barriers. The compiler part of the memory barriers
> > > > > effectively make _all_ variables temporily volatile. However, 2 of
> > > > > the
> > > > > the 4 accesses to this variable doesn't use an atomic op.
> > > > >
> > > > > % #ifdef SMP
> > > > > % /*
> > > > > % * We don't want multiple CPU's to panic at the same time, so we
> > > > > % * use panic_cpu as a simple spinlock. We have to keep checking
> > > > > % * panic_cpu if we are spinning in case the panic on the first
> > > > > % * CPU is canceled.
> > > > > % */
> > > > > % if (panic_cpu != PCPU_GET(cpuid))
> > > > >
> > > > > This access doesn't use an atomic op.
> > > > >
> > > > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > > > % PCPU_GET(cpuid)) == 0)
> > > > >
> > > > > This access uses an atomic op. Not all atomic ops use memory
> > > > > barriers,
> > > > > at least on i386. However, this one does, at least on i386.
> > > > >
> > > > > (I'm always confused about what atomic_any() without acq or release
> > > > > means. Do they mean that you don't want a memory barrier (this is
> > > > > what the mostly give, at least on i386), and if so, what use are
> > > > > they? There are far too many atomic ops, for far too many
> > > > > never-used
> > > > > widths, with alternative spellings to encourage using a wrong one.
> > > > > cmpset is is especially confusing since it you can spell it as
> > > > > cmpset,
> > > > > cmpset_acq or compset_rel and always get the barrier, at least on
> > > > > i386. At least cmpset only supports 1 width, at least on i386.)
> > > > >
> > > > > % while (panic_cpu != NOCPU)
> > > > >
> > > > > This access doesn't use an atomic op.
> > > > >
> > > > > % ; /* nothing */
> > > > > % #endif
> > > > > % ...
> > > > > % #ifdef RESTARTABLE_PANICS
> > > > > % /* See if the user aborted the panic, in which case we
> > > > > continue. */
> > > > > % if (panicstr == NULL) {
> > > > > % #ifdef SMP
> > > > > % atomic_store_rel_int(&panic_cpu, NOCPU);
> > > > >
> > > > > This access uses an atomic op with a memory barrier, at least on i386.
> > > > > Now its rel semantics are clear.
> > > > >
> > > > > panicstr is non-volatile and never accessed by an atomic op, so it
> > > > > probably
> > > > > strictly needs to be declared volatile even more than panic_cpu. I
> > > > > think
> > > >
> > > > I agree about panicstr.
> > > > But I am not sure if we have places in code (beyond panic function)
> > > > itself where
> > > > volatile vs. non-volatile would make any actual difference.
> > > > But still.
> > > >
> > > > > the only thing that makes it work now is that it is bogusly pubic, and
> > > > > compilers can't analyze the whole program yet -- if they could, then
> > > > > they
> > > > > would see that it is only set in panic().
> > > > >
> > > > > % #endif
> > > > > % return;
> > > > > % }
> > > > > % #endif
> > > > > % #endif
> > > > >
> > > > > Now, why don't the partial memory barriers prevent caching the
> > > > > variable?
> > > > >
> > > > > % if (panic_cpu != PCPU_GET(cpuid))
> > > > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > > > % PCPU_GET(cpuid)) == 0)
> > > > > % while (panic_cpu != NOCPU)
> > > > > % ; /* nothing */
> > > > >
> > > > > The very first access can't reasonably use a cachec value.
> > > > > atomic_cmpset()
> > > > > can change panic_cpu, so even without the m
Re: svn commit: r213779 - head/sys/dev/sound/pci
On Wednesday 13 October 2010 10:39 am, Rui Paulo wrote:
> Author: rpaulo
> Date: Wed Oct 13 14:39:54 2010
> New Revision: 213779
> URL: http://svn.freebsd.org/changeset/base/213779
>
> Log:
> Fix a brain-o: wrong case statement semantics.
>
> Found with: clang
>
> Modified:
> head/sys/dev/sound/pci/envy24ht.c
> head/sys/dev/sound/pci/spicds.c
>
> Modified: head/sys/dev/sound/pci/envy24ht.c
> ===
>=== --- head/sys/dev/sound/pci/envy24ht.c Wed Oct 13
> 14:37:52 2010 (r213778) +++ head/sys/dev/sound/pci/envy24ht.c Wed
> Oct 13 14:39:54 2010 (r213779) @@ -2236,7 +2236,8 @@
> envy24ht_putcfg(struct sc_info *sc)
> else
> printf("not implemented\n");
> switch (sc->adcn) {
> -case 0x01 || 0x02:
> +case 0x01:
> + case 0x02:
> printf(" ADC #: ");
> printf("%d\n", sc->adcn);
> break;
>
> Modified: head/sys/dev/sound/pci/spicds.c
> ===
>=== --- head/sys/dev/sound/pci/spicds.cWed Oct 13 14:37:52
> 2010 (r213778) +++ head/sys/dev/sound/pci/spicds.c Wed Oct 13
> 14:39:54 2010 (r213779) @@ -283,7 +283,8 @@ spicds_set(struct
> spicds_info *codec, in case SPICDS_TYPE_WM8770:
> left = left + 27;
> break;
> - case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:
> + case SPICDS_TYPE_AK4381:
> + case SPICDS_TYPE_AK4396:
> left = left * 255 / 100;
> break;
> default:
Although it was rediscovered by clang, spicds.c patch actually known
to fix a problem. Please see kern/146758.
Jung-uk Kim
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213798 - head/bin/sh
Author: obrien
Date: Wed Oct 13 18:23:43 2010
New Revision: 213798
URL: http://svn.freebsd.org/changeset/base/213798
Log:
If one wishes to set breakpoints of static the functions here, they
cannot be inlined.
Submitted by: jhb
Modified:
head/bin/sh/Makefile
Modified: head/bin/sh/Makefile
==
--- head/bin/sh/MakefileWed Oct 13 17:55:53 2010(r213797)
+++ head/bin/sh/MakefileWed Oct 13 18:23:43 2010(r213798)
@@ -21,7 +21,7 @@ LDADD= -ll -ledit -ltermcap
LFLAGS= -8 # 8-bit lex scanner for arithmetic
CFLAGS+=-DSHELL -I. -I${.CURDIR}
# for debug:
-# DEBUG_FLAGS+= -g -DDEBUG=3
+# DEBUG_FLAGS+= -g -DDEBUG=3 -fno-inline
WARNS?=2
WFORMAT=0
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213793 - in head/sys/dev: ce cp
On Wednesday 13 October 2010 01:27 pm, Roman Divacky wrote:
> On Wed, Oct 13, 2010 at 05:17:50PM +, Rui Paulo wrote:
> > Author: rpaulo
> > Date: Wed Oct 13 17:17:50 2010
> > New Revision: 213793
> > URL: http://svn.freebsd.org/changeset/base/213793
> >
> > Log:
> > Don't do a logical AND of the result of strcmp() with a
> > constant.
> >
> > Found with: clang
> >
> > Modified:
> > head/sys/dev/ce/if_ce.c
> > head/sys/dev/cp/if_cp.c
> >
> > Modified: head/sys/dev/ce/if_ce.c
> > =
> >= --- head/sys/dev/ce/if_ce.cWed Oct 13 17:16:08
> > 2010(r213792) +++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:17:50
> > 2010(r213793) @@ -1313,7 +1313,7 @@ static int ce_ioctl (struct
> > cdev *dev, u IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
> > IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
> > d->ifp->if_flags |= PP_CISCO;
> > - } else if (! strcmp ("fr", (char*)data) && PP_FR) {
> > + } else if (! strcmp ("fr", (char*)data)) {
>
> this is wrong I think... the PP_FR was used for compiling in/out
> support for something.. see the comment:
>
> /* If we don't have Cronyx's sppp version, we don't have fr support
> via sppp */ #ifndef PP_FR
> #define PP_FR 0
> #endif
>
> note that PP_FR is used in some other places as a flag. I guess
> that by compiling with something like make -DPP_FR=42 some magic
> would happen.
>
> anyway - this does not look like a bug but like an intent, please
> revert.
I think the attached patch should do.
Jung-uk Kim
Index: sys/dev/ce/if_ce.c
===
--- sys/dev/ce/if_ce.c (revision 213782)
+++ sys/dev/ce/if_ce.c (working copy)
@@ -1313,9 +1313,11 @@ static int ce_ioctl (struct cdev *dev, u_long cmd,
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data) && PP_FR) {
+#if PP_FR > 0
+ } else if (! strcmp ("fr", (char*)data)) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
+#endif
} else if (! strcmp ("ppp", (char*)data)) {
IFP2SP(d->ifp)->pp_flags &= ~PP_FR;
IFP2SP(d->ifp)->pp_flags &= ~PP_KEEPALIVE;
Index: sys/dev/cp/if_cp.c
===
--- sys/dev/cp/if_cp.c (revision 213782)
+++ sys/dev/cp/if_cp.c (working copy)
@@ -1052,9 +1052,11 @@ static int cp_ioctl (struct cdev *dev, u_long cmd,
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data) && PP_FR) {
+#if PP_FR > 0
+ } else if (! strcmp ("fr", (char*)data)) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
+#endif
} else if (! strcmp ("ppp", (char*)data)) {
IFP2SP(d->ifp)->pp_flags &= ~PP_FR;
IFP2SP(d->ifp)->pp_flags &= ~PP_KEEPALIVE;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213797 - head/sys/crypto/aesni
Author: dim Date: Wed Oct 13 17:55:53 2010 New Revision: 213797 URL: http://svn.freebsd.org/changeset/base/213797 Log: Change two missed instances of 'retq' in aeskeys_i386.S to 'retl', which makes it possible to assemble this file with gas from newer binutils. Reviewed by: kib Modified: head/sys/crypto/aesni/aeskeys_i386.S Modified: head/sys/crypto/aesni/aeskeys_i386.S == --- head/sys/crypto/aesni/aeskeys_i386.SWed Oct 13 17:55:19 2010 (r213796) +++ head/sys/crypto/aesni/aeskeys_i386.SWed Oct 13 17:55:53 2010 (r213797) @@ -52,7 +52,7 @@ _key_expansion_256a: pxor%xmm1,%xmm0 movaps %xmm0,(%edx) addl$0x10,%edx - retq + retl .cfi_endproc END(_key_expansion_128) @@ -76,7 +76,7 @@ ENTRY(_key_expansion_192a) shufps $0b01001110,%xmm2,%xmm1 movaps %xmm1,0x10(%edx) addl$0x20,%edx - retq + retl .cfi_endproc END(_key_expansion_192a) ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213796 - head/sys/pci
Author: yongari
Date: Wed Oct 13 17:55:19 2010
New Revision: 213796
URL: http://svn.freebsd.org/changeset/base/213796
Log:
Rewrite interrupt handler to give fairness for both RX and TX.
Previously rl(4) continuously checked whether there are RX events
or TX completions in forever loop. This caused TX starvation under
high RX load as well as consuming too much CPU cycles in the
interrupt handler. If interrupt was shared with other devices which
may be always true due to USB devices in these days, rl(4) also
tried to process the interrupt. This means polling(4) was the only
way to mitigate the these issues.
To address these issues, rl(4) now disables interrupts when it
knows the interrupt is ours and limit the number of iteration of
the loop to 16. The interrupt would be enabled again before exiting
interrupt handler if the driver is still running. Because RX buffer
is 64KB in size, the number of iterations in the loop has nothing
to do with number of RX packets being processed. This change
ensures sending TX frames under high RX load.
RX handler drops a driver lock to pass received frames to upper
stack such that there is a window that user can down the interface.
So rl(4) now checks whether driver is still running before serving
RX or TX completion in the loop.
While I'm here, exit interrupt handler when driver initialized
controller.
With this change, now rl(4) can send frames under high RX load even
though the TX performance is still not good(rl(4) controllers can't
queue more than 4 frames at a time so low TX performance was one of
design issue of rl(4) controllers). It's much better than previous
TX starvation and you should not notice RX performance drop with
this change. Controller still shows poor performance under high
network load but for many cases it's now usable without resorting
to polling(4).
MFC after:2 weeks
Modified:
head/sys/pci/if_rl.c
Modified: head/sys/pci/if_rl.c
==
--- head/sys/pci/if_rl.cWed Oct 13 17:38:23 2010(r213795)
+++ head/sys/pci/if_rl.cWed Oct 13 17:55:19 2010(r213796)
@@ -1620,6 +1620,7 @@ rl_intr(void *arg)
struct rl_softc *sc = arg;
struct ifnet*ifp = sc->rl_ifp;
uint16_tstatus;
+ int count;
RL_LOCK(sc);
@@ -1631,30 +1632,41 @@ rl_intr(void *arg)
goto done_locked;
#endif
- for (;;) {
+ if ((ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
+ goto done_locked2;
+ status = CSR_READ_2(sc, RL_ISR);
+ if (status == 0x || (status & RL_INTRS) == 0)
+ goto done_locked;
+ /*
+* Ours, disable further interrupts.
+*/
+ CSR_WRITE_2(sc, RL_IMR, 0);
+ for (count = 16; count > 0; count--) {
+ CSR_WRITE_2(sc, RL_ISR, status);
+ if (ifp->if_drv_flags & IFF_DRV_RUNNING) {
+ if (status & (RL_ISR_RX_OK | RL_ISR_RX_ERR))
+ rl_rxeof(sc);
+ if (status & (RL_ISR_TX_OK | RL_ISR_TX_ERR))
+ rl_txeof(sc);
+ if (status & RL_ISR_SYSTEM_ERR) {
+ ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
+ rl_init_locked(sc);
+ RL_UNLOCK(sc);
+ return;
+ }
+ }
status = CSR_READ_2(sc, RL_ISR);
/* If the card has gone away, the read returns 0x. */
- if (status == 0x)
+ if (status == 0x || (status & RL_INTRS) == 0)
break;
- if (status != 0)
- CSR_WRITE_2(sc, RL_ISR, status);
- if ((status & RL_INTRS) == 0)
- break;
- if (status & RL_ISR_RX_OK)
- rl_rxeof(sc);
- if (status & RL_ISR_RX_ERR)
- rl_rxeof(sc);
- if ((status & RL_ISR_TX_OK) || (status & RL_ISR_TX_ERR))
- rl_txeof(sc);
- if (status & RL_ISR_SYSTEM_ERR) {
- ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
- rl_init_locked(sc);
- }
}
if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
rl_start_locked(ifp);
+done_locked2:
+ if (ifp->if_drv_flags & IFF_DRV_RUNNING)
+ CSR_WRITE_2(sc, RL_IMR, RL_INTRS);
done_locked:
RL_UNLOCK(sc);
}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213795 - in head/sys/dev: ce cp
Author: rpaulo
Date: Wed Oct 13 17:38:23 2010
New Revision: 213795
URL: http://svn.freebsd.org/changeset/base/213795
Log:
Revert r213793.
Modified:
head/sys/dev/ce/if_ce.c
head/sys/dev/cp/if_cp.c
Modified: head/sys/dev/ce/if_ce.c
==
--- head/sys/dev/ce/if_ce.c Wed Oct 13 17:21:21 2010(r213794)
+++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:38:23 2010(r213795)
@@ -1313,7 +1313,7 @@ static int ce_ioctl (struct cdev *dev, u
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data)) {
+ } else if (! strcmp ("fr", (char*)data) && PP_FR) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
} else if (! strcmp ("ppp", (char*)data)) {
Modified: head/sys/dev/cp/if_cp.c
==
--- head/sys/dev/cp/if_cp.c Wed Oct 13 17:21:21 2010(r213794)
+++ head/sys/dev/cp/if_cp.c Wed Oct 13 17:38:23 2010(r213795)
@@ -1052,7 +1052,7 @@ static int cp_ioctl (struct cdev *dev, u
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data)) {
+ } else if (! strcmp ("fr", (char*)data) && PP_FR) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
} else if (! strcmp ("ppp", (char*)data)) {
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213793 - in head/sys/dev: ce cp
On Wed, Oct 13, 2010 at 05:17:50PM +, Rui Paulo wrote:
> Author: rpaulo
> Date: Wed Oct 13 17:17:50 2010
> New Revision: 213793
> URL: http://svn.freebsd.org/changeset/base/213793
>
> Log:
> Don't do a logical AND of the result of strcmp() with a constant.
>
> Found with: clang
>
> Modified:
> head/sys/dev/ce/if_ce.c
> head/sys/dev/cp/if_cp.c
>
> Modified: head/sys/dev/ce/if_ce.c
> ==
> --- head/sys/dev/ce/if_ce.c Wed Oct 13 17:16:08 2010(r213792)
> +++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:17:50 2010(r213793)
> @@ -1313,7 +1313,7 @@ static int ce_ioctl (struct cdev *dev, u
> IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
> IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
> d->ifp->if_flags |= PP_CISCO;
> - } else if (! strcmp ("fr", (char*)data) && PP_FR) {
> + } else if (! strcmp ("fr", (char*)data)) {
this is wrong I think... the PP_FR was used for compiling in/out
support for something.. see the comment:
/* If we don't have Cronyx's sppp version, we don't have fr support via sppp */
#ifndef PP_FR
#define PP_FR 0
#endif
note that PP_FR is used in some other places as a flag. I guess that by
compiling
with something like make -DPP_FR=42 some magic would happen.
anyway - this does not look like a bug but like an intent, please revert.
roman
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213794 - head/sys/netgraph
Author: rpaulo
Date: Wed Oct 13 17:21:21 2010
New Revision: 213794
URL: http://svn.freebsd.org/changeset/base/213794
Log:
When calling panic(), always pass a format string.
Modified:
head/sys/netgraph/ng_UI.c
head/sys/netgraph/ng_async.c
head/sys/netgraph/ng_frame_relay.c
head/sys/netgraph/ng_gif_demux.c
head/sys/netgraph/ng_iface.c
head/sys/netgraph/ng_rfc1490.c
head/sys/netgraph/ng_socket.c
head/sys/netgraph/ng_tty.c
Modified: head/sys/netgraph/ng_UI.c
==
--- head/sys/netgraph/ng_UI.c Wed Oct 13 17:17:50 2010(r213793)
+++ head/sys/netgraph/ng_UI.c Wed Oct 13 17:21:21 2010(r213794)
@@ -197,7 +197,7 @@ ng_UI_rcvdata(hook_p hook, item_p item)
mtod(m, u_char *)[0] = HDLC_UI;
NG_FWD_NEW_DATA(error, item, priv->downlink, m);/* m ->
NULL */
} else
- panic(__func__);
+ panic("%s", __func__);
done:
NG_FREE_M(m); /* does nothing if m == NULL */
@@ -234,7 +234,7 @@ ng_UI_disconnect(hook_p hook)
else if (hook == priv->uplink)
priv->uplink = NULL;
else
- panic(__func__);
+ panic("%s", __func__);
/*
* If we are not already shutting down,
* and we have no more hooks, then DO shut down.
Modified: head/sys/netgraph/ng_async.c
==
--- head/sys/netgraph/ng_async.cWed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_async.cWed Oct 13 17:21:21 2010
(r213794)
@@ -256,7 +256,7 @@ nga_rcvdata(hook_p hook, item_p item)
return (nga_rcv_sync(sc, item));
if (hook == sc->async)
return (nga_rcv_async(sc, item));
- panic(__func__);
+ panic("%s", __func__);
}
/*
@@ -372,7 +372,7 @@ nga_disconnect(hook_p hook)
else if (hook == sc->sync)
hookp = &sc->sync;
else
- panic(__func__);
+ panic("%s", __func__);
if (!*hookp)
panic("%s 2", __func__);
*hookp = NULL;
Modified: head/sys/netgraph/ng_frame_relay.c
==
--- head/sys/netgraph/ng_frame_relay.c Wed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_frame_relay.c Wed Oct 13 17:21:21 2010
(r213794)
@@ -396,7 +396,7 @@ ngfrm_rcvdata(hook_p hook, item_p item)
data[3] |= BYTEX_EA;
break;
default:
- panic(__func__);
+ panic("%s", __func__);
}
/* Send it */
Modified: head/sys/netgraph/ng_gif_demux.c
==
--- head/sys/netgraph/ng_gif_demux.cWed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_gif_demux.cWed Oct 13 17:21:21 2010
(r213794)
@@ -391,7 +391,7 @@ ng_gif_demux_disconnect(hook_p hook)
else {
iffam = get_iffam_from_hook(priv, hook);
if (iffam == NULL)
- panic(__func__);
+ panic("%s", __func__);
*get_hook_from_iffam(priv, iffam) = NULL;
}
Modified: head/sys/netgraph/ng_iface.c
==
--- head/sys/netgraph/ng_iface.cWed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_iface.cWed Oct 13 17:21:21 2010
(r213794)
@@ -821,7 +821,7 @@ ng_iface_disconnect(hook_p hook)
const iffam_p iffam = get_iffam_from_hook(priv, hook);
if (iffam == NULL)
- panic(__func__);
+ panic("%s", __func__);
*get_hook_from_iffam(priv, iffam) = NULL;
return (0);
}
Modified: head/sys/netgraph/ng_rfc1490.c
==
--- head/sys/netgraph/ng_rfc1490.c Wed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_rfc1490.c Wed Oct 13 17:21:21 2010
(r213794)
@@ -440,7 +440,7 @@ switch_on_etype:etype = ntohs(*((const
mtod(m, u_char *)[7] = 0x07;
NG_FWD_NEW_DATA(error, item, priv->downlink, m);
} else
- panic(__func__);
+ panic("%s", __func__);
done:
if (item)
@@ -485,7 +485,7 @@ ng_rfc1490_disconnect(hook_p hook)
else if (hook == priv->ethernet)
priv->ethernet = NULL;
else
- panic(__func__);
+ panic("%s", __func__);
return (0);
}
Modified: head/sys/netgraph/ng_socket.c
==
--- head/sys/netgraph/ng_socket.c Wed Oct 13 17:17:50 2010
(r213793)
+++ head/sys/netgraph/ng_socket.c
svn commit: r213793 - in head/sys/dev: ce cp
Author: rpaulo
Date: Wed Oct 13 17:17:50 2010
New Revision: 213793
URL: http://svn.freebsd.org/changeset/base/213793
Log:
Don't do a logical AND of the result of strcmp() with a constant.
Found with: clang
Modified:
head/sys/dev/ce/if_ce.c
head/sys/dev/cp/if_cp.c
Modified: head/sys/dev/ce/if_ce.c
==
--- head/sys/dev/ce/if_ce.c Wed Oct 13 17:16:08 2010(r213792)
+++ head/sys/dev/ce/if_ce.c Wed Oct 13 17:17:50 2010(r213793)
@@ -1313,7 +1313,7 @@ static int ce_ioctl (struct cdev *dev, u
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data) && PP_FR) {
+ } else if (! strcmp ("fr", (char*)data)) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
} else if (! strcmp ("ppp", (char*)data)) {
Modified: head/sys/dev/cp/if_cp.c
==
--- head/sys/dev/cp/if_cp.c Wed Oct 13 17:16:08 2010(r213792)
+++ head/sys/dev/cp/if_cp.c Wed Oct 13 17:17:50 2010(r213793)
@@ -1052,7 +1052,7 @@ static int cp_ioctl (struct cdev *dev, u
IFP2SP(d->ifp)->pp_flags &= ~(PP_FR);
IFP2SP(d->ifp)->pp_flags |= PP_KEEPALIVE;
d->ifp->if_flags |= PP_CISCO;
- } else if (! strcmp ("fr", (char*)data) && PP_FR) {
+ } else if (! strcmp ("fr", (char*)data)) {
d->ifp->if_flags &= ~(PP_CISCO);
IFP2SP(d->ifp)->pp_flags |= PP_FR | PP_KEEPALIVE;
} else if (! strcmp ("ppp", (char*)data)) {
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213792 - head/sys/contrib/pf/netinet
Author: rpaulo
Date: Wed Oct 13 17:16:08 2010
New Revision: 213792
URL: http://svn.freebsd.org/changeset/base/213792
Log:
Ignore the return value of ADDCARRY().
Modified:
head/sys/contrib/pf/netinet/in4_cksum.c
Modified: head/sys/contrib/pf/netinet/in4_cksum.c
==
--- head/sys/contrib/pf/netinet/in4_cksum.c Wed Oct 13 17:13:43 2010
(r213791)
+++ head/sys/contrib/pf/netinet/in4_cksum.c Wed Oct 13 17:16:08 2010
(r213792)
@@ -75,7 +75,7 @@
#include
#define ADDCARRY(x) (x > 65535 ? x -= 65535 : x)
-#define REDUCE {l_util.l = sum; sum = l_util.s[0] + l_util.s[1];
ADDCARRY(sum);}
+#define REDUCE {l_util.l = sum; sum = l_util.s[0] + l_util.s[1];
(void)ADDCARRY(sum);}
int in4_cksum(struct mbuf *, u_int8_t, int, int);
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213791 - head/sys/cddl/compat/opensolaris/kern
Author: rpaulo
Date: Wed Oct 13 17:13:43 2010
New Revision: 213791
URL: http://svn.freebsd.org/changeset/base/213791
Log:
Pass a format string to panic() and to taskqueue_start_threads().
Found with: clang
Modified:
head/sys/cddl/compat/opensolaris/kern/opensolaris_cmn_err.c
head/sys/cddl/compat/opensolaris/kern/opensolaris_taskq.c
Modified: head/sys/cddl/compat/opensolaris/kern/opensolaris_cmn_err.c
==
--- head/sys/cddl/compat/opensolaris/kern/opensolaris_cmn_err.c Wed Oct 13
17:12:23 2010(r213790)
+++ head/sys/cddl/compat/opensolaris/kern/opensolaris_cmn_err.c Wed Oct 13
17:13:43 2010(r213791)
@@ -48,7 +48,7 @@ vcmn_err(int ce, const char *fmt, va_lis
panic("Solaris: unknown severity level");
}
if (ce == CE_PANIC)
- panic(buf);
+ panic("%s", buf);
if (ce != CE_IGNORE)
vprintf(buf, adx);
}
Modified: head/sys/cddl/compat/opensolaris/kern/opensolaris_taskq.c
==
--- head/sys/cddl/compat/opensolaris/kern/opensolaris_taskq.c Wed Oct 13
17:12:23 2010(r213790)
+++ head/sys/cddl/compat/opensolaris/kern/opensolaris_taskq.c Wed Oct 13
17:13:43 2010(r213791)
@@ -73,7 +73,7 @@ taskq_create(const char *name, int nthre
tq = kmem_alloc(sizeof(*tq), KM_SLEEP);
tq->tq_queue = taskqueue_create(name, M_WAITOK,
taskqueue_thread_enqueue,
&tq->tq_queue);
- (void) taskqueue_start_threads(&tq->tq_queue, nthreads, pri, name);
+ (void) taskqueue_start_threads(&tq->tq_queue, nthreads, pri, "%s",
name);
return ((taskq_t *)tq);
}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213790 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: rpaulo Date: Wed Oct 13 17:12:23 2010 New Revision: 213790 URL: http://svn.freebsd.org/changeset/base/213790 Log: In zfs_post_common(), use %d instead of %hhu. Found with: clang Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_fm.c Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_fm.c == --- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_fm.cWed Oct 13 17:09:53 2010(r213789) +++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_fm.cWed Oct 13 17:12:23 2010(r213790) @@ -350,7 +350,7 @@ zfs_post_common(spa_t *spa, vdev_t *vd, snprintf(class, sizeof(class), "%s.%s.%s", FM_RSRC_RESOURCE, ZFS_ERROR_CLASS, name); - sbuf_printf(&sb, " %s=%hhu", FM_VERSION, FM_RSRC_VERSION); + sbuf_printf(&sb, " %s=%d", FM_VERSION, FM_RSRC_VERSION); sbuf_printf(&sb, " %s=%s", FM_CLASS, class); sbuf_printf(&sb, " %s=%ju", FM_EREPORT_PAYLOAD_ZFS_POOL_GUID, spa_guid(spa)); ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213789 - in head/sys/contrib/ngatm/netnatm: msg sig
Author: rpaulo
Date: Wed Oct 13 17:09:53 2010
New Revision: 213789
URL: http://svn.freebsd.org/changeset/base/213789
Log:
Properly tell the compiler that we want to ignore the return value of
certain macros.
Modified:
head/sys/contrib/ngatm/netnatm/msg/privmsg.c
head/sys/contrib/ngatm/netnatm/msg/uni_ie.c
head/sys/contrib/ngatm/netnatm/sig/sig_call.c
head/sys/contrib/ngatm/netnatm/sig/sig_reset.c
Modified: head/sys/contrib/ngatm/netnatm/msg/privmsg.c
==
--- head/sys/contrib/ngatm/netnatm/msg/privmsg.cWed Oct 13 17:09:16
2010(r213788)
+++ head/sys/contrib/ngatm/netnatm/msg/privmsg.cWed Oct 13 17:09:53
2010(r213789)
@@ -145,7 +145,7 @@ uni_decode_body_internal(enum uni_msgtyp
msg->b_rptr = msg->b_wptr;
else
msg->b_rptr += ielen;
- UNI_SAVE_IERR(cx, ietype, hdr.act, UNI_IERR_UNK);
+ (void)UNI_SAVE_IERR(cx, ietype, hdr.act, UNI_IERR_UNK);
err = -1;
continue;
}
@@ -200,16 +200,16 @@ uni_decode_body_internal(enum uni_msgtyp
* Unexpected but recognized.
* Q.2931 5.6.8.3
*/
- UNI_SAVE_IERR(cx, ietype, hdr.act, UNI_IERR_UNK);
+ (void)UNI_SAVE_IERR(cx, ietype, hdr.act, UNI_IERR_UNK);
err = -1;
break;
case DEC_ERR: /* bad IE */
if (iedecl->flags & UNIFL_ACCESS)
/* this may be wrong: 5.6.8.2 */
- UNI_SAVE_IERR(cx, ietype, hdr.act,
UNI_IERR_ACC);
+ (void)UNI_SAVE_IERR(cx, ietype, hdr.act,
UNI_IERR_ACC);
else
- UNI_SAVE_IERR(cx, ietype, hdr.act,
UNI_IERR_BAD);
+ (void)UNI_SAVE_IERR(cx, ietype, hdr.act,
UNI_IERR_BAD);
err = -1;
break;
Modified: head/sys/contrib/ngatm/netnatm/msg/uni_ie.c
==
--- head/sys/contrib/ngatm/netnatm/msg/uni_ie.c Wed Oct 13 17:09:16 2010
(r213788)
+++ head/sys/contrib/ngatm/netnatm/msg/uni_ie.c Wed Oct 13 17:09:53 2010
(r213789)
@@ -214,7 +214,7 @@ uni_encode_msg_hdr(struct uni_msg *msg,
{
u_char byte;
- uni_msg_ensure(msg, 9);
+ (void)uni_msg_ensure(msg, 9);
APP_BYTE(msg, cx->pnni ? PNNI_PROTO : UNI_PROTO);
APP_BYTE(msg, 3);
@@ -652,7 +652,7 @@ uni_encode_ie_hdr(struct uni_msg *msg, e
{
u_char byte;
- uni_msg_ensure(msg, 4 + len);
+ (void)uni_msg_ensure(msg, 4 + len);
*msg->b_wptr++ = type;
byte = 0x80 | (h->coding << 5);
Modified: head/sys/contrib/ngatm/netnatm/sig/sig_call.c
==
--- head/sys/contrib/ngatm/netnatm/sig/sig_call.c Wed Oct 13 17:09:16
2010(r213788)
+++ head/sys/contrib/ngatm/netnatm/sig/sig_call.c Wed Oct 13 17:09:53
2010(r213789)
@@ -398,7 +398,7 @@ un0_setup(struct call *c, struct uni_msg
if (IE_ISGOOD(u->u.setup.epref) &&
u->u.setup.epref.flag == 1) {
IE_SETERROR(u->u.setup.epref);
- UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
+ (void)UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
u->u.setup.epref.h.act, UNI_IERR_BAD);
}
uni_mandate_epref(c->uni, &u->u.setup.epref);
@@ -578,7 +578,7 @@ u1n6_call_proc(struct call *c, struct un
(cp->epref.flag != 1 ||
cp->epref.epref != c->msg_setup.epref.epref)) {
IE_SETERROR(cp->epref);
- UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
+ (void)UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
cp->epref.h.act, UNI_IERR_BAD);
}
}
@@ -763,7 +763,7 @@ unx_alerting(struct call *c, struct uni_
(al->epref.flag != 1 ||
al->epref.epref != c->msg_setup.epref.epref)) {
IE_SETERROR(al->epref);
- UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
+ (void)UNI_SAVE_IERR(&c->uni->cx, UNI_IE_EPREF,
al->epref.h.act, UNI_IERR_BAD);
}
}
@@ -1065,7 +1065,7 @@ unx_connect(struct call *c, struct uni_m
if (IE_ISGOOD(co->epref) &&
co->epref.flag != 1) {
IE_SETERROR(co->epref);
- UNI_SAVE_IE
svn commit: r213788 - head/sys/contrib/ngatm/netnatm/api
Author: rpaulo Date: Wed Oct 13 17:09:16 2010 New Revision: 213788 URL: http://svn.freebsd.org/changeset/base/213788 Log: Fix several cases were a conditional operator was used instead of a bitwise operator. Found with: clang Modified: head/sys/contrib/ngatm/netnatm/api/cc_conn.c Modified: head/sys/contrib/ngatm/netnatm/api/cc_conn.c == --- head/sys/contrib/ngatm/netnatm/api/cc_conn.cWed Oct 13 17:06:25 2010(r213787) +++ head/sys/contrib/ngatm/netnatm/api/cc_conn.cWed Oct 13 17:09:16 2010(r213788) @@ -977,25 +977,25 @@ cc_conn_sig_handle(struct ccconn *conn, /* * attributes */ - if (conn->dirty_attr && CCDIRTY_AAL) + if (conn->dirty_attr & CCDIRTY_AAL) resp->connect.aal = conn->aal; - if (conn->dirty_attr && CCDIRTY_BLLI) + if (conn->dirty_attr & CCDIRTY_BLLI) resp->connect.blli = conn->blli[conn->blli_selector - 1]; - if (conn->dirty_attr && CCDIRTY_CONNID) + if (conn->dirty_attr & CCDIRTY_CONNID) resp->connect.connid = conn->connid; /* XXX NOTIFY */ - if (conn->dirty_attr && CCDIRTY_EETD) + if (conn->dirty_attr & CCDIRTY_EETD) resp->connect.eetd = conn->eetd; /* XXX GIT */ /* XXX UU */ - if (conn->dirty_attr && CCDIRTY_TRAFFIC) + if (conn->dirty_attr & CCDIRTY_TRAFFIC) resp->connect.traffic = conn->traffic; - if (conn->dirty_attr && CCDIRTY_EXQOS) + if (conn->dirty_attr & CCDIRTY_EXQOS) resp->connect.exqos = conn->exqos; - if (conn->dirty_attr && CCDIRTY_ABRSETUP) + if (conn->dirty_attr & CCDIRTY_ABRSETUP) resp->connect.abrsetup = conn->abrsetup; - if (conn->dirty_attr && CCDIRTY_ABRADD) + if (conn->dirty_attr & CCDIRTY_ABRADD) resp->connect.abradd = conn->abradd; /* ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213787 - head/sys/dev/acpica/Osd
Author: jkim
Date: Wed Oct 13 17:06:25 2010
New Revision: 213787
URL: http://svn.freebsd.org/changeset/base/213787
Log:
Clean up unused headers.
Modified:
head/sys/dev/acpica/Osd/OsdHardware.c
Modified: head/sys/dev/acpica/Osd/OsdHardware.c
==
--- head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 17:01:33 2010
(r213786)
+++ head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 17:06:25 2010
(r213787)
@@ -34,12 +34,8 @@ __FBSDID("$FreeBSD$");
#include
-#include
-#include
#include
#include
-#include
-#include
/*
* ACPICA's rather gung-ho approach to hardware resource ownership is a little
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213786 - head/contrib/llvm/tools/clang/lib/Sema
Author: rdivacky Date: Wed Oct 13 17:01:33 2010 New Revision: 213786 URL: http://svn.freebsd.org/changeset/base/213786 Log: Actually, check for any kind of "C string type". Approved by:rpaulo (mentor) Modified: head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Modified: head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp == --- head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Wed Oct 13 16:57:06 2010(r213785) +++ head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Wed Oct 13 17:01:33 2010(r213786) @@ -1519,14 +1519,10 @@ CheckPrintfHandler::HandlePrintfSpecifie // Now type check the data expression that matches the // format specifier. const Expr *Ex = getDataArg(argIndex); -QualType Pointee = S.Context.UnsignedCharTy; -Pointee.addConst(); -QualType constType = (CS.getKind() == ConversionSpecifier::bArg) ? S.Context.IntTy : S.Context.getPointerType(Pointee); -QualType type = (CS.getKind() == ConversionSpecifier::bArg) ? S.Context.IntTy : S.Context.getPointerType(S.Context.UnsignedCharTy); -const analyze_printf::ArgTypeResult &ConstATR = constType; -const analyze_printf::ArgTypeResult &ATR = type; -if (ATR.isValid() && !ATR.matchesType(S.Context, Ex->getType()) && -!ConstATR.matchesType(S.Context, Ex->getType())) +const analyze_printf::ArgTypeResult &ATR = + (CS.getKind() == ConversionSpecifier::bArg) ? +ArgTypeResult(S.Context.IntTy) : ArgTypeResult::CStrTy; +if (ATR.isValid() && !ATR.matchesType(S.Context, Ex->getType())) S.Diag(getLocationOfByte(CS.getStart()), diag::warn_printf_conversion_argument_type_mismatch) << ATR.getRepresentativeType(S.Context) << Ex->getType() ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213785 - in head/lib/libc: net sys
Author: rpaulo Date: Wed Oct 13 16:57:06 2010 New Revision: 213785 URL: http://svn.freebsd.org/changeset/base/213785 Log: Clang related fixes: * When calling syslog(), pass a format string. * Define YY_NO_INPUT on nslexer.l Submitted by: Norberto Lopes Modified: head/lib/libc/net/nslexer.l head/lib/libc/sys/stack_protector.c Modified: head/lib/libc/net/nslexer.l == --- head/lib/libc/net/nslexer.l Wed Oct 13 16:34:08 2010(r213784) +++ head/lib/libc/net/nslexer.l Wed Oct 13 16:57:06 2010(r213785) @@ -53,6 +53,7 @@ static char *rcsid = #include "nsparser.h" +#defineYY_NO_INPUT #define YY_NO_UNPUT %} Modified: head/lib/libc/sys/stack_protector.c == --- head/lib/libc/sys/stack_protector.c Wed Oct 13 16:34:08 2010 (r213784) +++ head/lib/libc/sys/stack_protector.c Wed Oct 13 16:57:06 2010 (r213785) @@ -92,7 +92,7 @@ __fail(const char *msg) (void)sigprocmask(SIG_BLOCK, &mask, NULL); /* This may fail on a chroot jail... */ - syslog(LOG_CRIT, msg); + syslog(LOG_CRIT, "%s", msg); (void)memset(&sa, 0, sizeof(sa)); (void)sigemptyset(&sa.sa_mask); ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213784 - head/lib/libz
Author: imp
Date: Wed Oct 13 16:34:08 2010
New Revision: 213784
URL: http://svn.freebsd.org/changeset/base/213784
Log:
Revert 212517 to restore pristine state of this file
Modified:
head/lib/libz/minigzip.c
Modified: head/lib/libz/minigzip.c
==
--- head/lib/libz/minigzip.cWed Oct 13 16:30:41 2010(r213783)
+++ head/lib/libz/minigzip.cWed Oct 13 16:34:08 2010(r213784)
@@ -13,8 +13,6 @@
* or in pipe mode.
*/
-#include
-__FBSDID("$FreeBSD$");
/* @(#) $Id$ */
#include "zlib.h"
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213783 - head/sys/dev/acpica/Osd
Author: jkim
Date: Wed Oct 13 16:30:41 2010
New Revision: 213783
URL: http://svn.freebsd.org/changeset/base/213783
Log:
Remove acpi_bus_number() completely. It had to be removed in r212761.
Pointed out by: jhb
Modified:
head/sys/dev/acpica/Osd/OsdHardware.c
Modified: head/sys/dev/acpica/Osd/OsdHardware.c
==
--- head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 14:57:13 2010
(r213782)
+++ head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 16:30:41 2010
(r213783)
@@ -122,58 +122,3 @@ AcpiOsWritePciConfiguration (ACPI_PCI_ID
return (AE_OK);
}
-
-/*
- * Depth-first recursive case for finding the bus, given the slot/function.
- */
-static int __unused
-acpi_bus_number(ACPI_HANDLE root, ACPI_HANDLE curr, ACPI_PCI_ID *PciId)
-{
-ACPI_HANDLE parent;
-ACPI_STATUS status;
-ACPI_OBJECT_TYPE type;
-UINT32 adr;
-int bus, slot, func, class, subclass, header;
-
-/* Try to get the _BBN object of the root, otherwise assume it is 0. */
-bus = 0;
-if (root == curr) {
- status = acpi_GetInteger(root, "_BBN", &bus);
- if (ACPI_FAILURE(status) && bootverbose)
- printf("acpi_bus_number: root bus has no _BBN, assuming 0\n");
- return (bus);
-}
-status = AcpiGetParent(curr, &parent);
-if (ACPI_FAILURE(status))
- return (bus);
-
-/* First, recurse up the tree until we find the host bus. */
-bus = acpi_bus_number(root, parent, PciId);
-
-/* Validate parent bus device type. */
-if (ACPI_FAILURE(AcpiGetType(parent, &type)) || type != ACPI_TYPE_DEVICE) {
- printf("acpi_bus_number: not a device, type %d\n", type);
- return (bus);
-}
-
-/* Get the parent's slot and function. */
-status = acpi_GetInteger(parent, "_ADR", &adr);
-if (ACPI_FAILURE(status))
- return (bus);
-slot = ACPI_HIWORD(adr);
-func = ACPI_LOWORD(adr);
-
-/* Is this a PCI-PCI or Cardbus-PCI bridge? */
-class = pci_cfgregread(bus, slot, func, PCIR_CLASS, 1);
-if (class != PCIC_BRIDGE)
- return (bus);
-subclass = pci_cfgregread(bus, slot, func, PCIR_SUBCLASS, 1);
-
-/* Find the header type, masking off the multifunction bit. */
-header = pci_cfgregread(bus, slot, func, PCIR_HDRTYPE, 1) & PCIM_HDRTYPE;
-if (header == PCIM_HDRTYPE_BRIDGE && subclass == PCIS_BRIDGE_PCI)
- bus = pci_cfgregread(bus, slot, func, PCIR_SECBUS_1, 1);
-else if (header == PCIM_HDRTYPE_CARDBUS && subclass == PCIS_BRIDGE_CARDBUS)
- bus = pci_cfgregread(bus, slot, func, PCIR_SECBUS_2, 1);
-return (bus);
-}
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213772 - head/sys/dev/acpica/Osd
On Wednesday 13 October 2010 08:47 am, John Baldwin wrote: > On Wednesday, October 13, 2010 7:38:25 am Rui Paulo wrote: > > Author: rpaulo > > Date: Wed Oct 13 11:38:24 2010 > > New Revision: 213772 > > URL: http://svn.freebsd.org/changeset/base/213772 > > > > Log: > > Mark acpi_bus_number() as __unused. This allows clang to this > > file without any warnings. > > Actually, I think this should have been removed when > AcpiOsDerivePciId() was removed during the update to ACPICA > 20100915. Yes, you're right. I'll fix that. Thanks, Jung-uk Kim ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213648 - head/sys/kern
Thank you and Bruce for explanation.
The key point here (which I did not understand) is that
"something == PCPU_GET(cpuid))" may be true only if "something" is set by the
current cpu, in which case its value is not stale.
On Wed, Oct 13, 2010 at 09:05:04AM -0400, John Baldwin wrote:
> On Tuesday, October 12, 2010 5:08:04 pm Gennady Proskurin wrote:
> > On Sat, Oct 09, 2010 at 12:48:50PM +0300, Andriy Gapon wrote:
> > > on 09/10/2010 12:33 Bruce Evans said the following:
> > > > On Sat, 9 Oct 2010, Andriy Gapon wrote:
> > > >
> > > >> Log:
> > > >> panic_cpu variable should be volatile
> > > >>
> > > >> This is to prevent caching of its value in a register when it is
> > > >> checked
> > > >> and modified by multiple CPUs in parallel.
> > > >> Also, move the variable into the scope of the only function that
> > > >> uses it.
> > > >>
> > > >> Reviewed by:jhb
> > > >> Hint from:mdf
> > > >> MFC after:1 week
> > > >
> > > > I doubt that this is either necessary or sufficient. Most variables
> > > > aren't volatile while they are locked by a mutex. But panic() cannot
> > > > use
> > > > normal mutex locking, so it should access this variable using atomic
> > > > ops with memory barriers. The compiler part of the memory barriers
> > > > effectively make _all_ variables temporily volatile. However, 2 of the
> > > > the 4 accesses to this variable doesn't use an atomic op.
> > > >
> > > > % #ifdef SMP
> > > > % /*
> > > > % * We don't want multiple CPU's to panic at the same time, so we
> > > > % * use panic_cpu as a simple spinlock. We have to keep checking
> > > > % * panic_cpu if we are spinning in case the panic on the first
> > > > % * CPU is canceled.
> > > > % */
> > > > % if (panic_cpu != PCPU_GET(cpuid))
> > > >
> > > > This access doesn't use an atomic op.
> > > >
> > > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > > % PCPU_GET(cpuid)) == 0)
> > > >
> > > > This access uses an atomic op. Not all atomic ops use memory barriers,
> > > > at least on i386. However, this one does, at least on i386.
> > > >
> > > > (I'm always confused about what atomic_any() without acq or release
> > > > means. Do they mean that you don't want a memory barrier (this is
> > > > what the mostly give, at least on i386), and if so, what use are
> > > > they? There are far too many atomic ops, for far too many never-used
> > > > widths, with alternative spellings to encourage using a wrong one.
> > > > cmpset is is especially confusing since it you can spell it as cmpset,
> > > > cmpset_acq or compset_rel and always get the barrier, at least on
> > > > i386. At least cmpset only supports 1 width, at least on i386.)
> > > >
> > > > % while (panic_cpu != NOCPU)
> > > >
> > > > This access doesn't use an atomic op.
> > > >
> > > > % ; /* nothing */
> > > > % #endif
> > > > % ...
> > > > % #ifdef RESTARTABLE_PANICS
> > > > % /* See if the user aborted the panic, in which case we continue.
> > > > */
> > > > % if (panicstr == NULL) {
> > > > % #ifdef SMP
> > > > % atomic_store_rel_int(&panic_cpu, NOCPU);
> > > >
> > > > This access uses an atomic op with a memory barrier, at least on i386.
> > > > Now its rel semantics are clear.
> > > >
> > > > panicstr is non-volatile and never accessed by an atomic op, so it
> > > > probably
> > > > strictly needs to be declared volatile even more than panic_cpu. I
> > > > think
> > >
> > > I agree about panicstr.
> > > But I am not sure if we have places in code (beyond panic function)
> > > itself where
> > > volatile vs. non-volatile would make any actual difference.
> > > But still.
> > >
> > > > the only thing that makes it work now is that it is bogusly pubic, and
> > > > compilers can't analyze the whole program yet -- if they could, then
> > > > they
> > > > would see that it is only set in panic().
> > > >
> > > > % #endif
> > > > % return;
> > > > % }
> > > > % #endif
> > > > % #endif
> > > >
> > > > Now, why don't the partial memory barriers prevent caching the variable?
> > > >
> > > > % if (panic_cpu != PCPU_GET(cpuid))
> > > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > > % PCPU_GET(cpuid)) == 0)
> > > > % while (panic_cpu != NOCPU)
> > > > % ; /* nothing */
> > > >
> > > > The very first access can't reasonably use a cachec value.
> > > > atomic_cmpset()
> > > > can change panic_cpu, so even without the memory barrier, panic_cpu must
> > > > be reloaded for the third access the first time. But then when the
> > > > third
> > > > access is repeated in the second while loop, the missing atomic op with
> > > > barrier makes things very broken. panic_cpu isn't changed by the loop,
> > > > and the compiler thinks that it isn't changed by anything else either,
> > > > so
> > > > the compiler may reduce the loop to:
> > > >
> > > > %
svn commit: r213782 - head/sys/contrib/ipfilter/netinet
Author: rpaulo Date: Wed Oct 13 14:57:13 2010 New Revision: 213782 URL: http://svn.freebsd.org/changeset/base/213782 Log: Pass a format string to make_dev(). Found by: clang Modified: head/sys/contrib/ipfilter/netinet/mlfk_ipl.c Modified: head/sys/contrib/ipfilter/netinet/mlfk_ipl.c == --- head/sys/contrib/ipfilter/netinet/mlfk_ipl.cWed Oct 13 14:44:38 2010(r213781) +++ head/sys/contrib/ipfilter/netinet/mlfk_ipl.cWed Oct 13 14:57:13 2010(r213782) @@ -204,7 +204,7 @@ ipf_modload() } if (!c) c = str; - ipf_devs[i] = make_dev(&ipl_cdevsw, i, 0, 0, 0600, c); + ipf_devs[i] = make_dev(&ipl_cdevsw, i, 0, 0, 0600, "%s", c); } error = ipf_pfil_hook(); ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213781 - in head/sys/modules: cryptodev sysvipc/sysvmsg sysvipc/sysvsem
Author: rpaulo
Date: Wed Oct 13 14:44:38 2010
New Revision: 213781
URL: http://svn.freebsd.org/changeset/base/213781
Log:
Add opt_compat.h to SRCS.
Modified:
head/sys/modules/cryptodev/Makefile
head/sys/modules/sysvipc/sysvmsg/Makefile
head/sys/modules/sysvipc/sysvsem/Makefile
Modified: head/sys/modules/cryptodev/Makefile
==
--- head/sys/modules/cryptodev/Makefile Wed Oct 13 14:41:52 2010
(r213780)
+++ head/sys/modules/cryptodev/Makefile Wed Oct 13 14:44:38 2010
(r213781)
@@ -3,6 +3,6 @@
.PATH: ${.CURDIR}/../../opencrypto
KMOD = cryptodev
SRCS = cryptodev.c
-SRCS += bus_if.h device_if.h
+SRCS += bus_if.h device_if.h opt_compat.h
.include
Modified: head/sys/modules/sysvipc/sysvmsg/Makefile
==
--- head/sys/modules/sysvipc/sysvmsg/Makefile Wed Oct 13 14:41:52 2010
(r213780)
+++ head/sys/modules/sysvipc/sysvmsg/Makefile Wed Oct 13 14:44:38 2010
(r213781)
@@ -3,6 +3,6 @@
.PATH: ${.CURDIR}/../../../kern
KMOD= sysvmsg
-SRCS= sysv_msg.c opt_sysvipc.h
+SRCS= sysv_msg.c opt_sysvipc.h opt_compat.h
.include
Modified: head/sys/modules/sysvipc/sysvsem/Makefile
==
--- head/sys/modules/sysvipc/sysvsem/Makefile Wed Oct 13 14:41:52 2010
(r213780)
+++ head/sys/modules/sysvipc/sysvsem/Makefile Wed Oct 13 14:44:38 2010
(r213781)
@@ -3,6 +3,6 @@
.PATH: ${.CURDIR}/../../../kern
KMOD= sysvsem
-SRCS= sysv_sem.c opt_sysvipc.h
+SRCS= sysv_sem.c opt_sysvipc.h opt_compat.h
.include
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213780 - head/sys/dev/hptrr
Author: rpaulo Date: Wed Oct 13 14:41:52 2010 New Revision: 213780 URL: http://svn.freebsd.org/changeset/base/213780 Log: Pass a format string to make_dev(). Modified: head/sys/dev/hptrr/hptrr_osm_bsd.c Modified: head/sys/dev/hptrr/hptrr_osm_bsd.c == --- head/sys/dev/hptrr/hptrr_osm_bsd.c Wed Oct 13 14:39:54 2010 (r213779) +++ head/sys/dev/hptrr/hptrr_osm_bsd.c Wed Oct 13 14:41:52 2010 (r213780) @@ -1178,7 +1178,7 @@ static void hpt_final_init(void *dummy) } make_dev(&hpt_cdevsw, DRIVER_MINOR, UID_ROOT, GID_OPERATOR, - S_IRUSR | S_IWUSR, driver_name); + S_IRUSR | S_IWUSR, "%s", driver_name); } #if defined(KLD_MODULE) && (__FreeBSD_version >= 503000) ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213779 - head/sys/dev/sound/pci
Author: rpaulo
Date: Wed Oct 13 14:39:54 2010
New Revision: 213779
URL: http://svn.freebsd.org/changeset/base/213779
Log:
Fix a brain-o: wrong case statement semantics.
Found with: clang
Modified:
head/sys/dev/sound/pci/envy24ht.c
head/sys/dev/sound/pci/spicds.c
Modified: head/sys/dev/sound/pci/envy24ht.c
==
--- head/sys/dev/sound/pci/envy24ht.c Wed Oct 13 14:37:52 2010
(r213778)
+++ head/sys/dev/sound/pci/envy24ht.c Wed Oct 13 14:39:54 2010
(r213779)
@@ -2236,7 +2236,8 @@ envy24ht_putcfg(struct sc_info *sc)
else
printf("not implemented\n");
switch (sc->adcn) {
-case 0x01 || 0x02:
+case 0x01:
+ case 0x02:
printf(" ADC #: ");
printf("%d\n", sc->adcn);
break;
Modified: head/sys/dev/sound/pci/spicds.c
==
--- head/sys/dev/sound/pci/spicds.c Wed Oct 13 14:37:52 2010
(r213778)
+++ head/sys/dev/sound/pci/spicds.c Wed Oct 13 14:39:54 2010
(r213779)
@@ -283,7 +283,8 @@ spicds_set(struct spicds_info *codec, in
case SPICDS_TYPE_WM8770:
left = left + 27;
break;
- case SPICDS_TYPE_AK4381 || SPICDS_TYPE_AK4396:
+ case SPICDS_TYPE_AK4381:
+ case SPICDS_TYPE_AK4396:
left = left * 255 / 100;
break;
default:
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213778 - head/sys/dev/if_ndis
Author: rpaulo
Date: Wed Oct 13 14:37:52 2010
New Revision: 213778
URL: http://svn.freebsd.org/changeset/base/213778
Log:
WPA_CSE_WEP104 was being incorrectly checked.
Found with: clang
Modified:
head/sys/dev/if_ndis/if_ndis.c
Modified: head/sys/dev/if_ndis/if_ndis.c
==
--- head/sys/dev/if_ndis/if_ndis.c Wed Oct 13 14:27:47 2010
(r213777)
+++ head/sys/dev/if_ndis/if_ndis.c Wed Oct 13 14:37:52 2010
(r213778)
@@ -2115,7 +2115,7 @@ ndis_set_cipher(sc, cipher)
len = sizeof(arg);
- if (cipher == WPA_CSE_WEP40 || WPA_CSE_WEP104) {
+ if (cipher == WPA_CSE_WEP40 || cipher == WPA_CSE_WEP104) {
if (!(ic->ic_cryptocaps & IEEE80211_CRYPTO_WEP))
return (ENOTSUP);
arg = NDIS_80211_WEPSTAT_ENC1ENABLED;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213777 - head/contrib/llvm/tools/clang/lib/Sema
Author: rdivacky Date: Wed Oct 13 14:27:47 2010 New Revision: 213777 URL: http://svn.freebsd.org/changeset/base/213777 Log: Extend this check for const unsigned char *. Approved by:rpaulo (mentor) Modified: head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Modified: head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp == --- head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Wed Oct 13 14:02:45 2010(r213776) +++ head/contrib/llvm/tools/clang/lib/Sema/SemaChecking.cpp Wed Oct 13 14:27:47 2010(r213777) @@ -1519,10 +1519,14 @@ CheckPrintfHandler::HandlePrintfSpecifie // Now type check the data expression that matches the // format specifier. const Expr *Ex = getDataArg(argIndex); +QualType Pointee = S.Context.UnsignedCharTy; +Pointee.addConst(); +QualType constType = (CS.getKind() == ConversionSpecifier::bArg) ? S.Context.IntTy : S.Context.getPointerType(Pointee); QualType type = (CS.getKind() == ConversionSpecifier::bArg) ? S.Context.IntTy : S.Context.getPointerType(S.Context.UnsignedCharTy); -//const analyze_printf::ArgTypeResult &ATR = S.Context.IntTy; +const analyze_printf::ArgTypeResult &ConstATR = constType; const analyze_printf::ArgTypeResult &ATR = type; -if (ATR.isValid() && !ATR.matchesType(S.Context, Ex->getType())) +if (ATR.isValid() && !ATR.matchesType(S.Context, Ex->getType()) && +!ConstATR.matchesType(S.Context, Ex->getType())) S.Diag(getLocationOfByte(CS.getStart()), diag::warn_printf_conversion_argument_type_mismatch) << ATR.getRepresentativeType(S.Context) << Ex->getType() ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213744 - head/bin/sh
On Wednesday, October 13, 2010 12:29:27 am Bruce Evans wrote: > On Tue, 12 Oct 2010, David O'Brien wrote: > > > On Wed, Oct 13, 2010 at 02:18:33PM +1100, Bruce Evans wrote: > >> On Tue, 12 Oct 2010, David E. O'Brien wrote: > >>> Log: > >>> If DEBUG is 3 or greater, disable STATICization of functions. > >>> Also correct the documented location of the trace file. > >> > >> Private functions should always be static, which no `#define STATIC static' > > [..] > >> In theory, the debugging info should make it possible for debuggers > >> to restore the semantics of not-explictly-inline functions by virtualizing > >> them, but gdb's debugging info and/or gdb are too primitive to do this > >> (gdb doesn't allow putting a breakpoint at a deleted static function, > > > > This is actually what my motivation is -- trying to set breakpoints and > > finding GDB was unable to. > > > >> Of course, debugging and profiling are magic, > >> but I don't want to have to adorn all functions with STATICs and > >> __attributes() (and pragmas for othercc...) to recover historical/normal > >> or variant debugging or profiling of them. > > > > I agree, and would not add STATIC's to a program's code that didn't > > already have them. But in this case we inherited it from 4.4BSD. > > I'm just making it actually do something other than being a gratuitous > > spelling change. > > > > I believe I've made things more consistent with r213760. > > Add __noinline or whatever attributes to STATIC (but keep static in > it) for the DEBUG >= 3 case if you are going that far. __noinline > should be a syntax error for variables, so this should also find any > STATICs still on variables. The spelling fix of changing STATIC to > what it actually means > (ASSORTED_HACKS_FOR_DEBUGGING_BUT_NOW_ONLY_FOR_FUNCTIONS) goes too far > for me. To be honest, I think changing STATIC is excessive churn. The "right" fix is to use 'make DEBUG_FLAGS="-g -DDEBUG=2 -fno-inline"'. I often use 'make DEBUG_FLAGS="-g -fno-inline"' to workaround excessive inlining when debugging. I think all of the STATIC changes should just be backed out. -- John Baldwin ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213648 - head/sys/kern
On Tuesday, October 12, 2010 5:08:04 pm Gennady Proskurin wrote:
> On Sat, Oct 09, 2010 at 12:48:50PM +0300, Andriy Gapon wrote:
> > on 09/10/2010 12:33 Bruce Evans said the following:
> > > On Sat, 9 Oct 2010, Andriy Gapon wrote:
> > >
> > >> Log:
> > >> panic_cpu variable should be volatile
> > >>
> > >> This is to prevent caching of its value in a register when it is checked
> > >> and modified by multiple CPUs in parallel.
> > >> Also, move the variable into the scope of the only function that uses
> > >> it.
> > >>
> > >> Reviewed by:jhb
> > >> Hint from:mdf
> > >> MFC after:1 week
> > >
> > > I doubt that this is either necessary or sufficient. Most variables
> > > aren't volatile while they are locked by a mutex. But panic() cannot use
> > > normal mutex locking, so it should access this variable using atomic
> > > ops with memory barriers. The compiler part of the memory barriers
> > > effectively make _all_ variables temporily volatile. However, 2 of the
> > > the 4 accesses to this variable doesn't use an atomic op.
> > >
> > > % #ifdef SMP
> > > % /*
> > > % * We don't want multiple CPU's to panic at the same time, so we
> > > % * use panic_cpu as a simple spinlock. We have to keep checking
> > > % * panic_cpu if we are spinning in case the panic on the first
> > > % * CPU is canceled.
> > > % */
> > > % if (panic_cpu != PCPU_GET(cpuid))
> > >
> > > This access doesn't use an atomic op.
> > >
> > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > % PCPU_GET(cpuid)) == 0)
> > >
> > > This access uses an atomic op. Not all atomic ops use memory barriers,
> > > at least on i386. However, this one does, at least on i386.
> > >
> > > (I'm always confused about what atomic_any() without acq or release
> > > means. Do they mean that you don't want a memory barrier (this is
> > > what the mostly give, at least on i386), and if so, what use are
> > > they? There are far too many atomic ops, for far too many never-used
> > > widths, with alternative spellings to encourage using a wrong one.
> > > cmpset is is especially confusing since it you can spell it as cmpset,
> > > cmpset_acq or compset_rel and always get the barrier, at least on
> > > i386. At least cmpset only supports 1 width, at least on i386.)
> > >
> > > % while (panic_cpu != NOCPU)
> > >
> > > This access doesn't use an atomic op.
> > >
> > > % ; /* nothing */
> > > % #endif
> > > % ...
> > > % #ifdef RESTARTABLE_PANICS
> > > % /* See if the user aborted the panic, in which case we continue. */
> > > % if (panicstr == NULL) {
> > > % #ifdef SMP
> > > % atomic_store_rel_int(&panic_cpu, NOCPU);
> > >
> > > This access uses an atomic op with a memory barrier, at least on i386.
> > > Now its rel semantics are clear.
> > >
> > > panicstr is non-volatile and never accessed by an atomic op, so it
> > > probably
> > > strictly needs to be declared volatile even more than panic_cpu. I think
> >
> > I agree about panicstr.
> > But I am not sure if we have places in code (beyond panic function) itself
> > where
> > volatile vs. non-volatile would make any actual difference.
> > But still.
> >
> > > the only thing that makes it work now is that it is bogusly pubic, and
> > > compilers can't analyze the whole program yet -- if they could, then they
> > > would see that it is only set in panic().
> > >
> > > % #endif
> > > % return;
> > > % }
> > > % #endif
> > > % #endif
> > >
> > > Now, why don't the partial memory barriers prevent caching the variable?
> > >
> > > % if (panic_cpu != PCPU_GET(cpuid))
> > > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > > % PCPU_GET(cpuid)) == 0)
> > > % while (panic_cpu != NOCPU)
> > > % ; /* nothing */
> > >
> > > The very first access can't reasonably use a cachec value.
> > > atomic_cmpset()
> > > can change panic_cpu, so even without the memory barrier, panic_cpu must
> > > be reloaded for the third access the first time. But then when the third
> > > access is repeated in the second while loop, the missing atomic op with
> > > barrier makes things very broken. panic_cpu isn't changed by the loop,
> > > and the compiler thinks that it isn't changed by anything else either, so
> > > the compiler may reduce the loop to:
> > >
> > > % if (panic_cpu != NOCPU)
> > > % for (;;)
> > > % ; /* nothing */
> >
> >
> > Yes, it's exactly the last loop that had the problem.
> > On amd64 with -O2:
> > .loc 1 544 0
> > movlpanic_cpu(%rip), %eax
> > .LVL134:
> > .p2align 4,,7
> > .L210:
> > cmpl$255, %eax
> > jne .L210
> > jmp .L225
> >
> > > except I've seen claims that even an endless for loop can be optimized
> > > to nothing. Declaring panic_cpu as volatile prevents the compiler
svn commit: r213775 - head/bin/sh
Author: jhb
Date: Wed Oct 13 13:22:11 2010
New Revision: 213775
URL: http://svn.freebsd.org/changeset/base/213775
Log:
Make DEBUG traces 64-bit clean:
- Use %t to print ptrdiff_t values.
- Cast a ptrdiff_t value explicitly to int for a field width specifier.
While here, sort includes.
Submitted by: Garrett Cooper
Modified:
head/bin/sh/expand.c
head/bin/sh/jobs.c
Modified: head/bin/sh/expand.c
==
--- head/bin/sh/expand.cWed Oct 13 13:17:38 2010(r213774)
+++ head/bin/sh/expand.cWed Oct 13 13:22:11 2010(r213775)
@@ -43,14 +43,15 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
-#include
#include
-#include
-#include
-#include
+#include
+#include
#include
+#include
#include
+#include
#include
+#include
/*
* Routines to expand arguments to commands. We have to deal with
@@ -497,9 +498,9 @@ expbackq(union node *cmd, int quoted, in
exitstatus = waitforjob(in.jp, (int *)NULL);
if (quoted == 0)
recordregion(startloc, dest - stackblock(), 0);
- TRACE(("evalbackq: size=%d: \"%.*s\"\n",
- (dest - stackblock()) - startloc,
- (dest - stackblock()) - startloc,
+ TRACE(("expbackq: size=%td: \"%.*s\"\n",
+ ((dest - stackblock()) - startloc),
+ (int)((dest - stackblock()) - startloc),
stackblock() + startloc));
expdest = dest;
INTON;
Modified: head/bin/sh/jobs.c
==
--- head/bin/sh/jobs.c Wed Oct 13 13:17:38 2010(r213774)
+++ head/bin/sh/jobs.c Wed Oct 13 13:22:11 2010(r213775)
@@ -38,18 +38,18 @@ static char sccsid[] = "@(#)jobs.c 8.5 (
#include
__FBSDID("$FreeBSD$");
-#include
-#include
-#include
-#include
-#include
-#include
+#include
#include
-#include
-#include
#include
+#include
+#include
+#include
+#include
+#include
#include
-#include
+#include
+#include
+#include
#include "shell.h"
#if JOBS
@@ -680,7 +680,7 @@ makejob(union node *node __unused, int n
jp->ps = &jp->ps0;
}
INTON;
- TRACE(("makejob(%p, %d) returns %%%d\n", (void *)node, nprocs,
+ TRACE(("makejob(%p, %d) returns %%%td\n", (void *)node, nprocs,
jp - jobtab + 1));
return jp;
}
@@ -766,7 +766,7 @@ forkshell(struct job *jp, union node *n,
pid_t pid;
pid_t pgrp;
- TRACE(("forkshell(%%%d, %p, %d) called\n", jp - jobtab, (void *)n,
+ TRACE(("forkshell(%%%td, %p, %d) called\n", jp - jobtab, (void *)n,
mode));
INTOFF;
if (mode == FORK_BG)
@@ -903,7 +903,7 @@ waitforjob(struct job *jp, int *origstat
int st;
INTOFF;
- TRACE(("waitforjob(%%%d) called\n", jp - jobtab + 1));
+ TRACE(("waitforjob(%%%td) called\n", jp - jobtab + 1));
while (jp->state == 0)
if (dowait(1, jp) == -1)
dotrap();
@@ -1004,7 +1004,7 @@ dowait(int block, struct job *job)
if (stopped) { /* stopped or done */
int state = done? JOBDONE : JOBSTOPPED;
if (jp->state != state) {
- TRACE(("Job %d: changing state from %d
to %d\n", jp - jobtab + 1, jp->state, state));
+ TRACE(("Job %td: changing state from %d
to %d\n", jp - jobtab + 1, jp->state, state));
jp->state = state;
if (jp != job) {
if (done && !jp->remembered &&
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213774 - head/bin/sh
Author: jhb
Date: Wed Oct 13 13:17:38 2010
New Revision: 213774
URL: http://svn.freebsd.org/changeset/base/213774
Log:
Suggest that DEBUG_FLAGS be used to enable extra debugging rather than
frobbing CFLAGS directly. DEBUG_FLAGS is something that can be specified
on the make command line without having to edit the Makefile directly.
Submitted by: Garrett Cooper
Modified:
head/bin/sh/Makefile
Modified: head/bin/sh/Makefile
==
--- head/bin/sh/MakefileWed Oct 13 11:39:36 2010(r213773)
+++ head/bin/sh/MakefileWed Oct 13 13:17:38 2010(r213774)
@@ -21,7 +21,7 @@ LDADD= -ll -ledit -ltermcap
LFLAGS= -8 # 8-bit lex scanner for arithmetic
CFLAGS+=-DSHELL -I. -I${.CURDIR}
# for debug:
-# CFLAGS+= -g -DDEBUG=3
+# DEBUG_FLAGS+= -g -DDEBUG=3
WARNS?=2
WFORMAT=0
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213772 - head/sys/dev/acpica/Osd
On Wednesday, October 13, 2010 7:38:25 am Rui Paulo wrote: > Author: rpaulo > Date: Wed Oct 13 11:38:24 2010 > New Revision: 213772 > URL: http://svn.freebsd.org/changeset/base/213772 > > Log: > Mark acpi_bus_number() as __unused. This allows clang to this file > without any warnings. Actually, I think this should have been removed when AcpiOsDerivePciId() was removed during the update to ACPICA 20100915. -- John Baldwin ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213769 - head/sys/geom/part
On Wed, Oct 13, 2010 at 11:35:59AM +, Rui Paulo wrote:
> Author: rpaulo
> Date: Wed Oct 13 11:35:59 2010
> New Revision: 213769
> URL: http://svn.freebsd.org/changeset/base/213769
>
> Log:
> The canonical way to print __func__ when using KASSERT() is to write
> ("%s", __func__). This avoids clang's -Wformat-string warnings.
>
That's a stupid warning, clang should know the value of __func__.
Stefan
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213772 - head/sys/dev/acpica/Osd
Author: rpaulo
Date: Wed Oct 13 11:38:24 2010
New Revision: 213772
URL: http://svn.freebsd.org/changeset/base/213772
Log:
Mark acpi_bus_number() as __unused. This allows clang to this file
without any warnings.
Modified:
head/sys/dev/acpica/Osd/OsdHardware.c
Modified: head/sys/dev/acpica/Osd/OsdHardware.c
==
--- head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 11:37:39 2010
(r213771)
+++ head/sys/dev/acpica/Osd/OsdHardware.c Wed Oct 13 11:38:24 2010
(r213772)
@@ -126,7 +126,7 @@ AcpiOsWritePciConfiguration (ACPI_PCI_ID
/*
* Depth-first recursive case for finding the bus, given the slot/function.
*/
-static int
+static int __unused
acpi_bus_number(ACPI_HANDLE root, ACPI_HANDLE curr, ACPI_PCI_ID *PciId)
{
ACPI_HANDLE parent;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213771 - head/sys/fs/msdosfs
Author: rpaulo Date: Wed Oct 13 11:37:39 2010 New Revision: 213771 URL: http://svn.freebsd.org/changeset/base/213771 Log: Ignore the return value of DE_INTERNALIZE(). Modified: head/sys/fs/msdosfs/msdosfs_denode.c Modified: head/sys/fs/msdosfs/msdosfs_denode.c == --- head/sys/fs/msdosfs/msdosfs_denode.cWed Oct 13 11:37:12 2010 (r213770) +++ head/sys/fs/msdosfs/msdosfs_denode.cWed Oct 13 11:37:39 2010 (r213771) @@ -240,7 +240,7 @@ deget(pmp, dirclust, diroffset, depp) *depp = NULL; return (error); } - DE_INTERNALIZE(ldep, direntptr); + (void)DE_INTERNALIZE(ldep, direntptr); brelse(bp); } ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213770 - in head/sys/dev: kbd kbdmux syscons
Author: rpaulo
Date: Wed Oct 13 11:37:12 2010
New Revision: 213770
URL: http://svn.freebsd.org/changeset/base/213770
Log:
Explicitly tell the compiler that we don't care about the return value
of kbdd_ioctl().
Modified:
head/sys/dev/kbd/kbd.c
head/sys/dev/kbdmux/kbdmux.c
head/sys/dev/syscons/syscons.c
Modified: head/sys/dev/kbd/kbd.c
==
--- head/sys/dev/kbd/kbd.c Wed Oct 13 11:35:59 2010(r213769)
+++ head/sys/dev/kbd/kbd.c Wed Oct 13 11:37:12 2010(r213770)
@@ -224,7 +224,7 @@ kbd_register(keyboard_t *kbd)
strcpy(ki.kb_name, kbd->kb_name);
ki.kb_unit = kbd->kb_unit;
- kbdd_ioctl(mux, KBADDKBD, (caddr_t) &ki);
+ (void)kbdd_ioctl(mux, KBADDKBD, (caddr_t) &ki);
}
return (index);
@@ -241,7 +241,7 @@ kbd_register(keyboard_t *kbd)
strcpy(ki.kb_name, kbd->kb_name);
ki.kb_unit = kbd->kb_unit;
- kbdd_ioctl(mux, KBADDKBD, (caddr_t) &ki);
+ (void)kbdd_ioctl(mux, KBADDKBD, (caddr_t) &ki);
}
return (index);
@@ -1148,7 +1148,7 @@ genkbd_diag(keyboard_t *kbd, int level)
(s) |= l ## DOWN; \
(s) ^= l ## ED; \
i = (s) & LOCK_MASK;\
- kbdd_ioctl((k), KDSETLED, (caddr_t)&i); \
+ (void)kbdd_ioctl((k), KDSETLED, (caddr_t)&i); \
}
static u_int
@@ -1308,7 +1308,7 @@ genkbd_keyaction(keyboard_t *kbd, int ke
#else
state &= ~CLKED;
i = state & LOCK_MASK;
- kbdd_ioctl(kbd, KDSETLED, (caddr_t)&i);
+ (void)kbdd_ioctl(kbd, KDSETLED, (caddr_t)&i);
#endif
break;
case SLK:
@@ -1344,7 +1344,7 @@ genkbd_keyaction(keyboard_t *kbd, int ke
#else
state |= CLKED;
i = state & LOCK_MASK;
- kbdd_ioctl(kbd, KDSETLED, (caddr_t)&i);
+ (void)kbdd_ioctl(kbd, KDSETLED, (caddr_t)&i);
#endif
break;
case SLK:
Modified: head/sys/dev/kbdmux/kbdmux.c
==
--- head/sys/dev/kbdmux/kbdmux.cWed Oct 13 11:35:59 2010
(r213769)
+++ head/sys/dev/kbdmux/kbdmux.cWed Oct 13 11:37:12 2010
(r213770)
@@ -1115,7 +1115,7 @@ kbdmux_ioctl(keyboard_t *kbd, u_long cmd
/* KDSETLED on all slave keyboards */
SLIST_FOREACH(k, &state->ks_kbds, next)
- kbdd_ioctl(k->kbd, KDSETLED, arg);
+ (void)kbdd_ioctl(k->kbd, KDSETLED, arg);
KBDMUX_UNLOCK(state);
break;
@@ -1146,7 +1146,7 @@ kbdmux_ioctl(keyboard_t *kbd, u_long cmd
/* KDSKBSTATE on all slave keyboards */
SLIST_FOREACH(k, &state->ks_kbds, next)
- kbdd_ioctl(k->kbd, KDSKBSTATE, arg);
+ (void)kbdd_ioctl(k->kbd, KDSKBSTATE, arg);
KBDMUX_UNLOCK(state);
@@ -1192,7 +1192,7 @@ kbdmux_ioctl(keyboard_t *kbd, u_long cmd
/* perform command on all slave keyboards */
SLIST_FOREACH(k, &state->ks_kbds, next)
- kbdd_ioctl(k->kbd, cmd, arg);
+ (void)kbdd_ioctl(k->kbd, cmd, arg);
KBDMUX_UNLOCK(state);
break;
@@ -1205,7 +1205,7 @@ kbdmux_ioctl(keyboard_t *kbd, u_long cmd
/* perform command on all slave keyboards */
SLIST_FOREACH(k, &state->ks_kbds, next)
- kbdd_ioctl(k->kbd, cmd, arg);
+ (void)kbdd_ioctl(k->kbd, cmd, arg);
KBDMUX_UNLOCK(state);
/* FALLTHROUGH */
Modified: head/sys/dev/syscons/syscons.c
==
--- head/sys/dev/syscons/syscons.c Wed Oct 13 11:35:59 2010
(r213769)
+++ head/sys/dev/syscons/syscons.c Wed Oct 13 11:37:12 2010
(r213770)
@@ -475,7 +475,7 @@ sc_attach_unit(int unit, int flags)
scrn_timer(sc);
/* set up the keyboard */
-kbdd_ioctl(sc->kbd, KDSKBMODE, (caddr_t)&scp->kbd_mode);
+(void)kbdd_ioctl(sc->kbd, KDSKBMODE, (caddr_t)&scp->kbd_mode);
update_kbd_state(scp, scp->status, LOCK_MASK);
printf("%s%d: %s <%d virtual consoles, flags=0x%x>\n",
@@ -584,7 +584,7 @@ sctty_open(struct tty *tp)
#ifndef __sparc64__
if (s
svn commit: r213769 - head/sys/geom/part
Author: rpaulo
Date: Wed Oct 13 11:35:59 2010
New Revision: 213769
URL: http://svn.freebsd.org/changeset/base/213769
Log:
The canonical way to print __func__ when using KASSERT() is to write
("%s", __func__). This avoids clang's -Wformat-string warnings.
Modified:
head/sys/geom/part/g_part.c
head/sys/geom/part/g_part_ebr.c
head/sys/geom/part/g_part_mbr.c
Modified: head/sys/geom/part/g_part.c
==
--- head/sys/geom/part/g_part.c Wed Oct 13 11:23:27 2010(r213768)
+++ head/sys/geom/part/g_part.c Wed Oct 13 11:35:59 2010(r213769)
@@ -1597,7 +1597,8 @@ g_part_ctlreq(struct gctl_req *req, stru
(gpp.gpp_parms & G_PART_PARM_FLAGS) &&
strchr(gpp.gpp_flags, 'C') != NULL) ? 1 : 0;
if (auto_commit) {
- KASSERT(gpp.gpp_parms & G_PART_PARM_GEOM, (__func__));
+ KASSERT(gpp.gpp_parms & G_PART_PARM_GEOM, ("%s",
+ __func__));
error = g_part_ctl_commit(req, &gpp);
}
}
@@ -1737,11 +1738,11 @@ g_part_dumpconf(struct sbuf *sb, const c
struct g_part_entry *entry;
struct g_part_table *table;
- KASSERT(sb != NULL && gp != NULL, (__func__));
+ KASSERT(sb != NULL && gp != NULL, ("%s", __func__));
table = gp->softc;
if (indent == NULL) {
- KASSERT(cp == NULL && pp != NULL, (__func__));
+ KASSERT(cp == NULL && pp != NULL, ("%s", __func__));
entry = pp->private;
if (entry == NULL)
return;
@@ -1756,7 +1757,7 @@ g_part_dumpconf(struct sbuf *sb, const c
*/
G_PART_DUMPCONF(table, entry, sb, indent);
} else if (cp != NULL) {/* Consumer configuration. */
- KASSERT(pp == NULL, (__func__));
+ KASSERT(pp == NULL, ("%s", __func__));
/* none */
} else if (pp != NULL) {/* Provider configuration. */
entry = pp->private;
@@ -1799,11 +1800,11 @@ g_part_orphan(struct g_consumer *cp)
struct g_part_table *table;
pp = cp->provider;
- KASSERT(pp != NULL, (__func__));
+ KASSERT(pp != NULL, ("%s", __func__));
G_PART_TRACE((G_T_TOPOLOGY, "%s(%s)", __func__, pp->name));
g_topology_assert();
- KASSERT(pp->error != 0, (__func__));
+ KASSERT(pp->error != 0, ("%s", __func__));
table = cp->geom->softc;
if (table != NULL && table->gpt_opened)
g_access(cp, -1, -1, -1);
Modified: head/sys/geom/part/g_part_ebr.c
==
--- head/sys/geom/part/g_part_ebr.c Wed Oct 13 11:23:27 2010
(r213768)
+++ head/sys/geom/part/g_part_ebr.c Wed Oct 13 11:35:59 2010
(r213769)
@@ -221,8 +221,8 @@ g_part_ebr_add(struct g_part_table *base
if (baseentry->gpe_deleted)
bzero(&entry->ent, sizeof(entry->ent));
- KASSERT(baseentry->gpe_start <= start, (__func__));
- KASSERT(baseentry->gpe_end >= start + size - 1, (__func__));
+ KASSERT(baseentry->gpe_start <= start, ("%s", __func__));
+ KASSERT(baseentry->gpe_end >= start + size - 1, ("%s", __func__));
baseentry->gpe_index = (start / sectors) + 1;
baseentry->gpe_offset = (off_t)(start + sectors) * pp->sectorsize;
baseentry->gpe_start = start;
Modified: head/sys/geom/part/g_part_mbr.c
==
--- head/sys/geom/part/g_part_mbr.c Wed Oct 13 11:23:27 2010
(r213768)
+++ head/sys/geom/part/g_part_mbr.c Wed Oct 13 11:35:59 2010
(r213769)
@@ -204,8 +204,8 @@ g_part_mbr_add(struct g_part_table *base
if (baseentry->gpe_deleted)
bzero(&entry->ent, sizeof(entry->ent));
- KASSERT(baseentry->gpe_start <= start, (__func__));
- KASSERT(baseentry->gpe_end >= start + size - 1, (__func__));
+ KASSERT(baseentry->gpe_start <= start, ("%s", __func__));
+ KASSERT(baseentry->gpe_end >= start + size - 1, ("%s", __func__));
baseentry->gpe_start = start;
baseentry->gpe_end = start + size - 1;
entry->ent.dp_start = start;
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213766 - head/sys/netinet6
Author: rpaulo
Date: Wed Oct 13 10:45:22 2010
New Revision: 213766
URL: http://svn.freebsd.org/changeset/base/213766
Log:
Purposely tell the compiler that we ignore the return value of ADDCARRY()
in the REDUCE macro.
Reviewed by: dim, rdivacky
Modified:
head/sys/netinet6/in6_cksum.c
Modified: head/sys/netinet6/in6_cksum.c
==
--- head/sys/netinet6/in6_cksum.c Wed Oct 13 10:33:01 2010
(r213765)
+++ head/sys/netinet6/in6_cksum.c Wed Oct 13 10:45:22 2010
(r213766)
@@ -78,7 +78,7 @@ __FBSDID("$FreeBSD$");
*/
#define ADDCARRY(x) (x > 65535 ? x -= 65535 : x)
-#define REDUCE {l_util.l = sum; sum = l_util.s[0] + l_util.s[1];
ADDCARRY(sum);}
+#define REDUCE {l_util.l = sum; sum = l_util.s[0] + l_util.s[1];
(void)ADDCARRY(sum);}
/*
* m MUST contain a continuous IP6 header.
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r213648 - head/sys/kern
On Sat, Oct 09, 2010 at 12:48:50PM +0300, Andriy Gapon wrote:
> on 09/10/2010 12:33 Bruce Evans said the following:
> > On Sat, 9 Oct 2010, Andriy Gapon wrote:
> >
> >> Log:
> >> panic_cpu variable should be volatile
> >>
> >> This is to prevent caching of its value in a register when it is checked
> >> and modified by multiple CPUs in parallel.
> >> Also, move the variable into the scope of the only function that uses it.
> >>
> >> Reviewed by:jhb
> >> Hint from:mdf
> >> MFC after:1 week
> >
> > I doubt that this is either necessary or sufficient. Most variables
> > aren't volatile while they are locked by a mutex. But panic() cannot use
> > normal mutex locking, so it should access this variable using atomic
> > ops with memory barriers. The compiler part of the memory barriers
> > effectively make _all_ variables temporily volatile. However, 2 of the
> > the 4 accesses to this variable doesn't use an atomic op.
> >
> > % #ifdef SMP
> > % /*
> > % * We don't want multiple CPU's to panic at the same time, so we
> > % * use panic_cpu as a simple spinlock. We have to keep checking
> > % * panic_cpu if we are spinning in case the panic on the first
> > % * CPU is canceled.
> > % */
> > % if (panic_cpu != PCPU_GET(cpuid))
> >
> > This access doesn't use an atomic op.
> >
> > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > % PCPU_GET(cpuid)) == 0)
> >
> > This access uses an atomic op. Not all atomic ops use memory barriers,
> > at least on i386. However, this one does, at least on i386.
> >
> > (I'm always confused about what atomic_any() without acq or release
> > means. Do they mean that you don't want a memory barrier (this is
> > what the mostly give, at least on i386), and if so, what use are
> > they? There are far too many atomic ops, for far too many never-used
> > widths, with alternative spellings to encourage using a wrong one.
> > cmpset is is especially confusing since it you can spell it as cmpset,
> > cmpset_acq or compset_rel and always get the barrier, at least on
> > i386. At least cmpset only supports 1 width, at least on i386.)
> >
> > % while (panic_cpu != NOCPU)
> >
> > This access doesn't use an atomic op.
> >
> > % ; /* nothing */
> > % #endif
> > % ...
> > % #ifdef RESTARTABLE_PANICS
> > % /* See if the user aborted the panic, in which case we continue. */
> > % if (panicstr == NULL) {
> > % #ifdef SMP
> > % atomic_store_rel_int(&panic_cpu, NOCPU);
> >
> > This access uses an atomic op with a memory barrier, at least on i386.
> > Now its rel semantics are clear.
> >
> > panicstr is non-volatile and never accessed by an atomic op, so it probably
> > strictly needs to be declared volatile even more than panic_cpu. I think
>
> I agree about panicstr.
> But I am not sure if we have places in code (beyond panic function) itself
> where
> volatile vs. non-volatile would make any actual difference.
> But still.
>
> > the only thing that makes it work now is that it is bogusly pubic, and
> > compilers can't analyze the whole program yet -- if they could, then they
> > would see that it is only set in panic().
> >
> > % #endif
> > % return;
> > % }
> > % #endif
> > % #endif
> >
> > Now, why don't the partial memory barriers prevent caching the variable?
> >
> > % if (panic_cpu != PCPU_GET(cpuid))
> > % while (atomic_cmpset_int(&panic_cpu, NOCPU,
> > % PCPU_GET(cpuid)) == 0)
> > % while (panic_cpu != NOCPU)
> > % ; /* nothing */
> >
> > The very first access can't reasonably use a cachec value. atomic_cmpset()
> > can change panic_cpu, so even without the memory barrier, panic_cpu must
> > be reloaded for the third access the first time. But then when the third
> > access is repeated in the second while loop, the missing atomic op with
> > barrier makes things very broken. panic_cpu isn't changed by the loop,
> > and the compiler thinks that it isn't changed by anything else either, so
> > the compiler may reduce the loop to:
> >
> > % if (panic_cpu != NOCPU)
> > % for (;;)
> > % ; /* nothing */
>
>
> Yes, it's exactly the last loop that had the problem.
> On amd64 with -O2:
> .loc 1 544 0
> movlpanic_cpu(%rip), %eax
> .LVL134:
> .p2align 4,,7
> .L210:
> cmpl$255, %eax
> jne .L210
> jmp .L225
>
> > except I've seen claims that even an endless for loop can be optimized
> > to nothing. Declaring panic_cpu as volatile prevents the compiler doing
> > this, but I think it is insufficient. We really do want to see panic_cpu
> > changed by other CPUs, and what is the point of atomic_load_acq*() if not
> > to use for this -- if declaring things volatile were sufficient, then we
> > could just use *(volatile *)&var. atomic_load/store_acq/rel*() on i386
>
> I
svn commit: r213765 - head/sys/dev/aic7xxx/aicasm
Author: rpaulo Date: Wed Oct 13 10:33:01 2010 New Revision: 213765 URL: http://svn.freebsd.org/changeset/base/213765 Log: Define YY_NO_INPUT. This makes aicasm buildable by clang with Werror turned on. Modified: head/sys/dev/aic7xxx/aicasm/aicasm_macro_scan.l head/sys/dev/aic7xxx/aicasm/aicasm_scan.l Modified: head/sys/dev/aic7xxx/aicasm/aicasm_macro_scan.l == --- head/sys/dev/aic7xxx/aicasm/aicasm_macro_scan.l Wed Oct 13 10:31:32 2010(r213764) +++ head/sys/dev/aic7xxx/aicasm/aicasm_macro_scan.l Wed Oct 13 10:33:01 2010(r213765) @@ -61,6 +61,7 @@ #include "aicasm_symbol.h" #include "aicasm_macro_gram.h" +#define YY_NO_INPUT #define MAX_STR_CONST 4096 static char string_buf[MAX_STR_CONST]; static char *string_buf_ptr; Modified: head/sys/dev/aic7xxx/aicasm/aicasm_scan.l == --- head/sys/dev/aic7xxx/aicasm/aicasm_scan.l Wed Oct 13 10:31:32 2010 (r213764) +++ head/sys/dev/aic7xxx/aicasm/aicasm_scan.l Wed Oct 13 10:33:01 2010 (r213765) @@ -61,6 +61,7 @@ #include "aicasm_symbol.h" #include "aicasm_gram.h" +#define YY_NO_INPUT /* This is used for macro body capture too, so err on the large size. */ #define MAX_STR_CONST 4096 static char string_buf[MAX_STR_CONST]; ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213764 - head/usr.bin/lex
Author: rpaulo Date: Wed Oct 13 10:31:32 2010 New Revision: 213764 URL: http://svn.freebsd.org/changeset/base/213764 Log: Don't define the input() function ifdef YY_NO_INPUT. This was previously done for the input() function. Submitted by: Norberto Lopes Modified: head/usr.bin/lex/flex.skl Modified: head/usr.bin/lex/flex.skl == --- head/usr.bin/lex/flex.skl Wed Oct 13 09:33:26 2010(r213763) +++ head/usr.bin/lex/flex.skl Wed Oct 13 10:31:32 2010(r213764) @@ -979,6 +979,7 @@ void yyFlexLexer::yyunput( int c, char* %- +#ifndef YY_NO_INPUT #ifdef __cplusplus static int yyinput() #else @@ -1054,6 +1055,7 @@ int yyFlexLexer::yyinput() return c; } +#endif /* ifndef YY_NO_INPUT */ %- ___ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213763 - head/usr.sbin/sysinstall
Author: brucec
Date: Wed Oct 13 09:33:26 2010
New Revision: 213763
URL: http://svn.freebsd.org/changeset/base/213763
Log:
Use the RFC2606 domain example.com in examples.
Modified:
head/usr.sbin/sysinstall/tcpip.c
Modified: head/usr.sbin/sysinstall/tcpip.c
==
--- head/usr.sbin/sysinstall/tcpip.cWed Oct 13 09:17:44 2010
(r213762)
+++ head/usr.sbin/sysinstall/tcpip.cWed Oct 13 09:33:26 2010
(r213763)
@@ -72,12 +72,12 @@ static char ipv6addr[INET6_ADDRSTRLEN];
static Layout layout[] = {
#define LAYOUT_HOSTNAME0
{ 1, 2, 25, HOSTNAME_FIELD_LEN - 1,
- "Host:", "Your fully-qualified hostname, e.g. foo.bar.com",
+ "Host:", "Your fully-qualified hostname, e.g. foo.example.com",
hostname, STRINGOBJ, NULL },
#define LAYOUT_DOMAINNAME 1
{ 1, 35, 20, HOSTNAME_FIELD_LEN - 1,
"Domain:",
- "The name of the domain that your machine is in, e.g. bar.com",
+ "The name of the domain that your machine is in, e.g. example.com",
domainname, STRINGOBJ, NULL },
#define LAYOUT_GATEWAY 2
{ 5, 2, 18, IPADDR_FIELD_LEN - 1,
___
svn-src-head@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r213762 - in head/sys/mips: cavium cavium/octe conf
Author: jmallett
Date: Wed Oct 13 09:17:44 2010
New Revision: 213762
URL: http://svn.freebsd.org/changeset/base/213762
Log:
o) Make it possible to attach a PHY directly to an octe device rather than
using miibus, since for some devices that use multiple addresses on the
bus,
going through miibus may be unclear, and for devices that are not standard
MII PHYs, miibus may throw a fit, necessitating complicated interfaces to
fake the interface that it expects during probe/attach.
o) Make the mv88e61xx SMI interface in octe attach a PHY directly and fix some
mistakes in the code that resulted from trying too hard to present a nice
interface to miibus.
o) Add a PHY driver for the mv88e61xx. If attached (it is optional in kernel
compiles so the default behavior of having a dumb switch is preserved) it
will place the switch in a VLAN-tagging mode such that each physical port
has a VLAN associated with it and interfaces for the VLANs can be created
to
address or bridge between them.
XXX It would be nice for this to be part of a single module including the
SMI interface, and for it to fit into a generic switch configuration
framework and for it to use DSA rather than VLANs, but this is a start
and gives some sense of the parameters of such frameworks that are not
currently present in FreeBSD. In lieu of a switch configuration
interface, per-port media status and VLAN settings are in a sysctl
tree.
XXX There may be some minor nits remaining in the handling of broadcast,
multicast and unknown destination traffic. It would also be nice to go
through and replace the few remaining magic numbers with macros at some
point in the future.
XXX This has only been tested with the MV88E6161, but it should work with
minimal or no modification on related switches, so support for probing
them was included.
Thanks to Pat Saavedra of TELoIP and Rafal Jaworowski of Semihalf for their
assistance in understanding the switch chipset.
Added:
head/sys/mips/cavium/octe/mv88e61xxphy.c (contents, props changed)
head/sys/mips/cavium/octe/mv88e61xxphyreg.h (contents, props changed)
Modified:
head/sys/mips/cavium/files.octeon1
head/sys/mips/cavium/octe/cavium-ethernet.h
head/sys/mips/cavium/octe/ethernet-mdio.c
head/sys/mips/cavium/octe/ethernet-mv88e61xx.c
head/sys/mips/cavium/octe/octe.c
head/sys/mips/conf/OCTEON1
Modified: head/sys/mips/cavium/files.octeon1
==
--- head/sys/mips/cavium/files.octeon1 Wed Oct 13 06:28:40 2010
(r213761)
+++ head/sys/mips/cavium/files.octeon1 Wed Oct 13 09:17:44 2010
(r213762)
@@ -31,6 +31,7 @@ mips/cavium/octe/ethernet-sgmii.c optio
mips/cavium/octe/ethernet-spi.coptional octe
mips/cavium/octe/ethernet-tx.c optional octe
mips/cavium/octe/ethernet-xaui.c optional octe
+mips/cavium/octe/mv88e61xxphy.coptional octe
mv88e61xxphy
mips/cavium/octe/octe.coptional octe
mips/cavium/octe/octebus.c optional octe
Modified: head/sys/mips/cavium/octe/cavium-ethernet.h
==
--- head/sys/mips/cavium/octe/cavium-ethernet.h Wed Oct 13 06:28:40 2010
(r213761)
+++ head/sys/mips/cavium/octe/cavium-ethernet.h Wed Oct 13 09:17:44 2010
(r213762)
@@ -72,6 +72,7 @@ typedef struct {
uint8_t mac[6];
int phy_id;
+ const char *phy_device;
int (*mdio_read)(struct ifnet *, int, int);
void (*mdio_write)(struct ifnet *, int, int, int);
Modified: head/sys/mips/cavium/octe/ethernet-mdio.c
==
--- head/sys/mips/cavium/octe/ethernet-mdio.c Wed Oct 13 06:28:40 2010
(r213761)
+++ head/sys/mips/cavium/octe/ethernet-mdio.c Wed Oct 13 09:17:44 2010
(r213762)
@@ -132,6 +132,7 @@ int cvm_oct_mdio_setup_device(struct ifn
cvm_oct_private_t *priv = (cvm_oct_private_t *)ifp->if_softc;
priv->phy_id = cvmx_helper_board_get_mii_address(priv->port);
+ priv->phy_device = NULL;
priv->mdio_read = NULL;
priv->mdio_write = NULL;
Modified: head/sys/mips/cavium/octe/ethernet-mv88e61xx.c
==
--- head/sys/mips/cavium/octe/ethernet-mv88e61xx.c Wed Oct 13 06:28:40
2010(r213761)
+++ head/sys/mips/cavium/octe/ethernet-mv88e61xx.c Wed Oct 13 09:17:44
2010(r213762)
@@ -49,8 +49,6 @@ __FBSDID("$FreeBSD$");
#include "wrapper-cvmx-includes.h"
#include "ethernet-headers.h"
-#defineMV88E61XX_SMI_PHY_SW0x10/* Switch PHY. */
-
#defineMV88E61XX_SMI_REG_CMD
