svn commit: r206648 - head/sys/cam/scsi
Author: avg
Date: Thu Apr 15 08:29:14 2010
New Revision: 206648
URL: http://svn.freebsd.org/changeset/base/206648
Log:
scsi_cd: CD_FLAG_VALID_MEDIA is sufficient to set d_sectorsize and
d_mediasize
CD_FLAG_VALID_TOC is not required for setting those media properties.
PR: kern/145385
Submitted by: Juergen Lock
a slightly different version
Tested by:Pavel Sukhoy ,
Markus Wild ,
Juergen Lock ,
uqs
MFC after:1 week
Modified:
head/sys/cam/scsi/scsi_cd.c
Modified: head/sys/cam/scsi/scsi_cd.c
==
--- head/sys/cam/scsi/scsi_cd.c Thu Apr 15 08:20:57 2010(r206647)
+++ head/sys/cam/scsi/scsi_cd.c Thu Apr 15 08:29:14 2010(r206648)
@@ -2773,8 +2773,12 @@ cdcheckmedia(struct cam_periph *periph)
softc->flags &= ~(CD_FLAG_VALID_MEDIA|CD_FLAG_VALID_TOC);
cdprevent(periph, PR_ALLOW);
return (error);
- } else
+ } else {
softc->flags |= CD_FLAG_VALID_MEDIA;
+ softc->disk->d_sectorsize = softc->params.blksize;
+ softc->disk->d_mediasize =
+ (off_t)softc->params.blksize * softc->params.disksize;
+ }
/*
* Now we check the table of contents. This (currently) is only
@@ -2863,9 +2867,6 @@ cdcheckmedia(struct cam_periph *periph)
}
softc->flags |= CD_FLAG_VALID_TOC;
- softc->disk->d_sectorsize = softc->params.blksize;
- softc->disk->d_mediasize =
- (off_t)softc->params.blksize * softc->params.disksize;
bailout:
___
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: r206649 - head/lib/libc/sys
Author: kib Date: Thu Apr 15 08:32:50 2010 New Revision: 206649 URL: http://svn.freebsd.org/changeset/base/206649 Log: Still reference struct __sigaction with clarification when this form of argument declaration is needed. Discussed with: bde MFC after:3 days Modified: head/lib/libc/sys/sigaction.2 Modified: head/lib/libc/sys/sigaction.2 == --- head/lib/libc/sys/sigaction.2 Thu Apr 15 08:29:14 2010 (r206648) +++ head/lib/libc/sys/sigaction.2 Thu Apr 15 08:32:50 2010 (r206649) @@ -496,6 +496,16 @@ or .Dv SIG_IGN this way. .Pp +If preprocessing symbol +.Va _POSIX_C_SOURCE +with the value >= 199309 is not defined, the following declaration for +the handler shall be used: +.Bl -tag -offset indent -width short +.It Tn POSIX Dv SA_SIGINFO : +.Ft void +.Fn handler int "struct __sigaction *" "void *" ; +.El +.Pp If the .Dv SA_SIGINFO flag is not set, the handler function should match ___ 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: r206650 - head/sys/geom
Author: avg Date: Thu Apr 15 08:39:56 2010 New Revision: 206650 URL: http://svn.freebsd.org/changeset/base/206650 Log: g_io_check: respond to zero pp->mediasize with ENXIO Previsouly this condition was reported with EIO by bio_offset > mediasize check. Perhaps that check should be extended to bio_offset+bio_length > mediasize. MFC after:1 week Modified: head/sys/geom/geom_io.c Modified: head/sys/geom/geom_io.c == --- head/sys/geom/geom_io.c Thu Apr 15 08:32:50 2010(r206649) +++ head/sys/geom/geom_io.c Thu Apr 15 08:39:56 2010(r206650) @@ -309,8 +309,8 @@ g_io_check(struct bio *bp) case BIO_READ: case BIO_WRITE: case BIO_DELETE: - /* Zero sectorsize is a probably lack of media */ - if (pp->sectorsize == 0) + /* Zero sectorsize or mediasize is probably a lack of media. */ + if (pp->sectorsize == 0 || pp->mediasize == 0) return (ENXIO); /* Reject I/O not on sector boundary */ if (bp->bio_offset % pp->sectorsize) ___ 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: r206651 - head/sys/cam/scsi
Author: avg Date: Thu Apr 15 09:22:14 2010 New Revision: 206651 URL: http://svn.freebsd.org/changeset/base/206651 Log: scsi_cd: CD_FLAG_VALID_MEDIA is sufficient to set d_sectorsize and d_mediasize [Forced commit to correct PR number.] CD_FLAG_VALID_TOC is not required for setting those media properties. PR: kern/138789 Submitted by: Juergen Lock a slightly different version Tested by:Pavel Sukhoy , Markus Wild , Juergen Lock , uqs MFC after:1 week Modified: head/sys/cam/scsi/scsi_cd.c Modified: head/sys/cam/scsi/scsi_cd.c == ___ 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: r206652 - head/sys/dev/siis
Author: mav Date: Thu Apr 15 11:17:33 2010 New Revision: 206652 URL: http://svn.freebsd.org/changeset/base/206652 Log: Explicitly enable PCI busmastering on attach. Now SiI3124 with siis(4) successfully works on sparc64 (SunBlade 100). H/W donated by: Gheorghe Ardelean Modified: head/sys/dev/siis/siis.c Modified: head/sys/dev/siis/siis.c == --- head/sys/dev/siis/siis.cThu Apr 15 09:22:14 2010(r206651) +++ head/sys/dev/siis/siis.cThu Apr 15 11:17:33 2010(r206652) @@ -164,6 +164,7 @@ siis_attach(device_t dev) rman_fini(&ctlr->sc_iomem); return (error); } + pci_enable_busmaster(dev); /* Reset controller */ siis_resume(dev); /* Number of HW channels */ ___ 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: r206659 - head/sys/netipsec
Author: vanhu
Date: Thu Apr 15 12:40:33 2010
New Revision: 206659
URL: http://svn.freebsd.org/changeset/base/206659
Log:
Locks SPTREE when setting some SP entries to state DEAD.
This can prevent kernel panics when updating SPs while
there is some traffic for them.
Obtained from: NETASQ
MFC after: 1m
Modified:
head/sys/netipsec/key.c
Modified: head/sys/netipsec/key.c
==
--- head/sys/netipsec/key.c Thu Apr 15 12:40:02 2010(r206658)
+++ head/sys/netipsec/key.c Thu Apr 15 12:40:33 2010(r206659)
@@ -1882,7 +1882,9 @@ key_spdadd(so, m, mhp)
newsp = key_getsp(&spidx);
if (mhp->msg->sadb_msg_type == SADB_X_SPDUPDATE) {
if (newsp) {
+ SPTREE_LOCK();
newsp->state = IPSEC_SPSTATE_DEAD;
+ SPTREE_UNLOCK();
KEY_FREESP(&newsp);
}
} else {
@@ -2117,7 +2119,9 @@ key_spddelete(so, m, mhp)
/* save policy id to buffer to be returned. */
xpl0->sadb_x_policy_id = sp->id;
+ SPTREE_LOCK();
sp->state = IPSEC_SPSTATE_DEAD;
+ SPTREE_UNLOCK();
KEY_FREESP(&sp);
{
@@ -2184,7 +2188,9 @@ key_spddelete2(so, m, mhp)
return key_senderror(so, m, EINVAL);
}
+ SPTREE_LOCK();
sp->state = IPSEC_SPSTATE_DEAD;
+ SPTREE_UNLOCK();
KEY_FREESP(&sp);
{
___
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: r206560 - head/sys/fs/devfs
On 2010-04-14, Kostik Belousov wrote: > First, I think we shall sort dirent->dd_dlist so that symlinks are > located before any other entries, and allow for duplication of > any name with the name of symlink. What will happen when user deletes a symlink covering a device? Uncovering the device may a bit surprising? If user created directories will be allowed one day, they will face the same problem. > I remember there was a PR to change devfs_symlink() so that symlink is > added to dd_dlist into head, not to a tail, as it is done now. kern/114057 > Assume that all full devfs pathes except symlinks are stored somewhere > with type information for the component at the end. > Then, just iterating this set at make_dev() time, we can verify that > name is not a duplicate, and that the path components are not already > used as device name (only as directories). > The pathes storage need to be protected by dev_mtx. > > Possibly, the cdev_priv.c_dev.__si_namebuf may be reused at the > cost of long check at the make_dev() time. Could you elaborate on this? I am not sure if I understand what you mean with __si_namebuf reuse. > What do you think of this ? I don't see anything obviously wrong with it. -- Jaakko ___ 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: r206560 - head/sys/fs/devfs
On Thu, Apr 15, 2010 at 04:01:56PM +0300, Jaakko Heinonen wrote: > On 2010-04-14, Kostik Belousov wrote: > > First, I think we shall sort dirent->dd_dlist so that symlinks are > > located before any other entries, and allow for duplication of > > any name with the name of symlink. > > What will happen when user deletes a symlink covering a device? > Uncovering the device may a bit surprising? If user created directories > will be allowed one day, they will face the same problem. Yes, device will be uncovered if it is there, the same with the directory. Use case I have in mind when I wrote this is the following: assume we have devices named dev1 and dev2, and two jails, j1 and j2, each with its own devfs mount. Also, assume that some program A has hardcoded /dev/dev1 as the name of device to operate, and it is run both in j1 and j2. Further, j1 is given access to /dev/dev1 only, while j2 is given access to /dev/dev2 only. Then administrator of j1 can create a symlink /dev/dev -> dev1, while administrator of j2 symlinks /dev/dev ->dev2, and both j1 and j2 would operate properly. This is not that perversed as it sounds, I think it is an essence of symlink usage on devfs. > > > I remember there was a PR to change devfs_symlink() so that symlink is > > added to dd_dlist into head, not to a tail, as it is done now. > > kern/114057 > > > Assume that all full devfs pathes except symlinks are stored somewhere > > with type information for the component at the end. > > Then, just iterating this set at make_dev() time, we can verify that > > name is not a duplicate, and that the path components are not already > > used as device name (only as directories). > > The pathes storage need to be protected by dev_mtx. > > > > Possibly, the cdev_priv.c_dev.__si_namebuf may be reused at the > > cost of long check at the make_dev() time. > > Could you elaborate on this? I am not sure if I understand what you mean > with __si_namebuf reuse. I mean that we do not to create the structure that mirrors the directory/nodes hierarchy for devfs dirents, but can dynamically parse __si_namebuf at the make_dev() time, iterating over all nodes using cdevp_list. For symlinks, it seems the easiest route is to have separate ref-counted structures containing the pathes for owning directories. > > > What do you think of this ? > > I don't see anything obviously wrong with it. > > -- > Jaakko pgpctzhUaU8zL.pgp Description: PGP signature
svn commit: r206662 - head/sys/dev/mxge
Author: gallatin
Date: Thu Apr 15 13:50:55 2010
New Revision: 206662
URL: http://svn.freebsd.org/changeset/base/206662
Log:
Cleanup if_media handling in mxge(4)
- Re-probe xfp / sfp+ socket on link events, in case user
has changed transceiver
- correctly report current media to avoid confusing lagg (reported by Panasas)
- Report link speed (submitted by yongari)
Reviewed by: yongari (earlier version)
MFC after: 7 days
Modified:
head/sys/dev/mxge/if_mxge.c
head/sys/dev/mxge/if_mxge_var.h
Modified: head/sys/dev/mxge/if_mxge.c
==
--- head/sys/dev/mxge/if_mxge.c Thu Apr 15 12:46:16 2010(r206661)
+++ head/sys/dev/mxge/if_mxge.c Thu Apr 15 13:50:55 2010(r206662)
@@ -883,6 +883,9 @@ mxge_send_cmd(mxge_softc_t *sc, uint32_t
case MXGEFW_CMD_ERROR_BUSY:
err = EBUSY;
break;
+ case MXGEFW_CMD_ERROR_I2C_ABSENT:
+ err = ENXIO;
+ break;
default:
device_printf(sc->dev,
"mxge: command %d "
@@ -2782,37 +2785,25 @@ static struct mxge_media_type mxge_sfp_m
};
static void
-mxge_set_media(mxge_softc_t *sc, int type)
+mxge_media_set(mxge_softc_t *sc, int media_type)
{
- sc->media_flags |= type;
- ifmedia_add(&sc->media, sc->media_flags, 0, NULL);
- ifmedia_set(&sc->media, sc->media_flags);
-}
+
+ ifmedia_add(&sc->media, IFM_ETHER | IFM_FDX | media_type,
+ 0, NULL);
+ ifmedia_set(&sc->media, IFM_ETHER | IFM_FDX | media_type);
+ sc->current_media = media_type;
+ sc->media.ifm_media = sc->media.ifm_cur->ifm_media;
+}
-/*
- * Determine the media type for a NIC. Some XFPs will identify
- * themselves only when their link is up, so this is initiated via a
- * link up interrupt. However, this can potentially take up to
- * several milliseconds, so it is run via the watchdog routine, rather
- * than in the interrupt handler itself. This need only be done
- * once, not each time the link is up.
- */
static void
-mxge_media_probe(mxge_softc_t *sc)
+mxge_media_init(mxge_softc_t *sc)
{
- mxge_cmd_t cmd;
- char *cage_type;
char *ptr;
- struct mxge_media_type *mxge_media_types = NULL;
- int i, err, ms, mxge_media_type_entries;
- uint32_t byte;
-
- sc->need_media_probe = 0;
+ int i;
- /* if we've already set a media type, we're done */
- if (sc->media_flags != (IFM_ETHER | IFM_AUTO))
- return;
+ ifmedia_removeall(&sc->media);
+ mxge_media_set(sc, IFM_AUTO);
/*
* parse the product code to deterimine the interface type
@@ -2823,6 +2814,7 @@ mxge_media_probe(mxge_softc_t *sc)
ptr = sc->product_code_string;
if (ptr == NULL) {
device_printf(sc->dev, "Missing product code\n");
+ return;
}
for (i = 0; i < 3; i++, ptr++) {
@@ -2835,17 +2827,44 @@ mxge_media_probe(mxge_softc_t *sc)
}
if (*ptr == 'C') {
/* -C is CX4 */
- mxge_set_media(sc, IFM_10G_CX4);
- return;
- }
- else if (*ptr == 'Q') {
+ sc->connector = MXGE_CX4;
+ mxge_media_set(sc, IFM_10G_CX4);
+ } else if (*ptr == 'Q') {
/* -Q is Quad Ribbon Fiber */
+ sc->connector = MXGE_QRF;
device_printf(sc->dev, "Quad Ribbon Fiber Media\n");
/* FreeBSD has no media type for Quad ribbon fiber */
- return;
+ } else if (*ptr == 'R') {
+ /* -R is XFP */
+ sc->connector = MXGE_XFP;
+ } else if (*ptr == 'S' || *(ptr +1) == 'S') {
+ /* -S or -2S is SFP+ */
+ sc->connector = MXGE_SFP;
+ } else {
+ device_printf(sc->dev, "Unknown media type: %c\n", *ptr);
}
+}
- if (*ptr == 'R') {
+/*
+ * Determine the media type for a NIC. Some XFPs will identify
+ * themselves only when their link is up, so this is initiated via a
+ * link up interrupt. However, this can potentially take up to
+ * several milliseconds, so it is run via the watchdog routine, rather
+ * than in the interrupt handler itself.
+ */
+static void
+mxge_media_probe(mxge_softc_t *sc)
+{
+ mxge_cmd_t cmd;
+ char *cage_type;
+
+ struct mxge_media_type *mxge_media_types = NULL;
+ int i, err, ms, mxge_media_type_entries;
+ uint32_t byte;
+
+ sc->need_media_probe = 0;
+
+ if (sc->connector == MXGE_XFP) {
/* -R is XFP */
mxge_media_types = mxge_xfp_media_types;
mxge_media_type_entries =
@@ -2853,9 +2872,7 @@ mxge_media_probe(mxge_softc_t *sc)
sizeof (mxge_xfp_media_types[0]);
svn commit: r206663 - head/sys/dev/mxge
Author: gallatin Date: Thu Apr 15 14:26:52 2010 New Revision: 206663 URL: http://svn.freebsd.org/changeset/base/206663 Log: Add missing IFCAP_LINKSTATE to mxge Submitted by: yongari Modified: head/sys/dev/mxge/if_mxge.c Modified: head/sys/dev/mxge/if_mxge.c == --- head/sys/dev/mxge/if_mxge.c Thu Apr 15 13:50:55 2010(r206662) +++ head/sys/dev/mxge/if_mxge.c Thu Apr 15 14:26:52 2010(r206663) @@ -4759,7 +4759,7 @@ mxge_attach(device_t dev) ifp->if_baudrate = IF_Gbps(10UL); ifp->if_capabilities = IFCAP_RXCSUM | IFCAP_TXCSUM | IFCAP_TSO4 | - IFCAP_VLAN_MTU; + IFCAP_VLAN_MTU | IFCAP_LINKSTATE; #ifdef INET ifp->if_capabilities |= IFCAP_LRO; #endif ___ 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: r206497 - in head: sbin/geom/class sbin/geom/class/sched sys/geom/sched sys/modules/geom sys/modules/geom/geom_sched sys/modules/geom/geom_sched/gs_sched sys/modules/geom/geom_sched/gs
On 04/12/2010 06:37 PM, Luigi Rizzo wrote: Bring in geom_sched, support for scheduling disk I/O requests in a device independent manner. Also include an example anticipatory scheduler, gsched_rr, which gives very nice performance improvements in presence of competing random access patterns. This is joint work with Fabio Checconi, developed last year and presented at BSDCan 2009. You can find details in the README file or at Are there any plans to bring in something like Linux's ionice (or even better solutions)? Is it possible to do it easily at all (the user and GEOM are seemingly very far from each other)? ___ 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: r206497 - in head: sbin/geom/class sbin/geom/class/sched sys/geom/sched sys/modules/geom sys/modules/geom/geom_sched sys/modules/geom/geom_sched/gs_sched sys/modules/geom/geom_sched/gs
On Thu, Apr 15, 2010 at 04:38:23PM +0200, Attila Nagy wrote: > On 04/12/2010 06:37 PM, Luigi Rizzo wrote: > > Bring in geom_sched, support for scheduling disk I/O requests > > in a device independent manner. Also include an example anticipatory > > scheduler, gsched_rr, which gives very nice performance improvements > > in presence of competing random access patterns. > > > > This is joint work with Fabio Checconi, developed last year > > and presented at BSDCan 2009. You can find details in the > > README file or at > > > Are there any plans to bring in something like Linux's ionice (or even > better solutions)? Is it possible to do it easily at all (the user and > GEOM are seemingly very far from each other)? I can easily[*] modify the I/O priority/weight of process or threads, but you have to keep in mind that once a request is intercepted by the filesystem code (e.g. for delayed writes, or avoid duplication of pending reads), by the time it hits the disk it is not trivial to tell who should be charged for it. This is an area where we still need to experiment a bit (especially to find on which process issues which request, say for metadata, delayed writes, shared reads). [*] the 'easy' way is put the information in an external hash table and look it up using as a search key either of uid, gid, . I think we did something like this in the GSOC2005 project on disk scheduling. A more intrusive way involves extending the struct thread with an extra field to store the io priority. cheers luigi ___ 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: r206664 - head/usr.sbin/config
Author: imp
Date: Thu Apr 15 15:10:46 2010
New Revision: 206664
URL: http://svn.freebsd.org/changeset/base/206664
Log:
Allow option aliasing. Lines of the form:
OLD_OPT = NEW_OPT
in options* files will now map OLD_OPT to NEW_OPT with a friendly
message. This is indented for situations where we need to preserve an
interface in the config file in an upwards compatible fashion on a
stable branch.
Reviewed by: nwhitehorn@
MFC after:3 days
Modified:
head/usr.sbin/config/config.h
head/usr.sbin/config/mkoptions.c
Modified: head/usr.sbin/config/config.h
==
--- head/usr.sbin/config/config.h Thu Apr 15 14:26:52 2010
(r206663)
+++ head/usr.sbin/config/config.h Thu Apr 15 15:10:46 2010
(r206664)
@@ -129,6 +129,8 @@ SLIST_HEAD(opt_head, opt) opt, mkopt, rm
struct opt_list {
char *o_name;
char *o_file;
+ int o_flags;
+#define OL_ALIAS 1
SLIST_ENTRY(opt_list) o_next;
};
Modified: head/usr.sbin/config/mkoptions.c
==
--- head/usr.sbin/config/mkoptions.cThu Apr 15 14:26:52 2010
(r206663)
+++ head/usr.sbin/config/mkoptions.cThu Apr 15 15:10:46 2010
(r206664)
@@ -94,6 +94,17 @@ options(void)
SLIST_INSERT_HEAD(&opt, op, op_next);
read_options();
+ SLIST_FOREACH(op, &opt, op_next) {
+ SLIST_FOREACH(ol, &otab, o_next) {
+ if (eq(op->op_name, ol->o_name) &&
+ (ol->o_flags & OL_ALIAS)) {
+ printf("Mapping option %s to %s.\n",
+ op->op_name, ol->o_file);
+ op->op_name = ol->o_file;
+ break;
+ }
+ }
+ }
SLIST_FOREACH(ol, &otab, o_next)
do_option(ol->o_name);
SLIST_FOREACH(op, &opt, op_next) {
@@ -124,7 +135,6 @@ do_option(char *name)
int tidy;
file = tooption(name);
-
/*
* Check to see if the option was specified..
*/
@@ -292,6 +302,7 @@ read_options(void)
struct opt_list *po;
int first = 1;
char genopt[MAXPATHLEN];
+ int flags = 0;
SLIST_INIT(&otab);
(void) snprintf(fname, sizeof(fname), "../../conf/options");
@@ -301,6 +312,7 @@ openit:
return;
}
next:
+ flags = 0;
wd = get_word(fp);
if (wd == (char *)EOF) {
(void) fclose(fp);
@@ -332,6 +344,18 @@ next:
(void) snprintf(genopt, sizeof(genopt), "opt_%s.h", lower(s));
val = genopt;
free(s);
+ } else if (eq(val, "=")) {
+ val = get_word(fp);
+ if (val == (char *)EOF) {
+ printf("%s: unexpected end of file\n", fname);
+ exit(1);
+ }
+ if (val == 0) {
+ printf("%s: Expected a right hand side at %s\n", fname,
+ this);
+ exit(1);
+ }
+ flags |= OL_ALIAS;
}
val = ns(val);
@@ -348,6 +372,7 @@ next:
err(EXIT_FAILURE, "calloc");
po->o_name = this;
po->o_file = val;
+ po->o_flags = flags;
SLIST_INSERT_HEAD(&otab, po, o_next);
goto next;
___
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: r206665 - head/sys/geom/eli
Author: pjd Date: Thu Apr 15 16:34:06 2010 New Revision: 206665 URL: http://svn.freebsd.org/changeset/base/206665 Log: Use lower priority for GELI worker threads. This improves system responsiveness under heavy GELI load. MFC after:3 days Modified: head/sys/geom/eli/g_eli.c Modified: head/sys/geom/eli/g_eli.c == --- head/sys/geom/eli/g_eli.c Thu Apr 15 15:10:46 2010(r206664) +++ head/sys/geom/eli/g_eli.c Thu Apr 15 16:34:06 2010(r206665) @@ -340,7 +340,7 @@ g_eli_worker(void *arg) } #endif thread_lock(curthread); - sched_prio(curthread, PRIBIO); + sched_prio(curthread, PUSER); if (sc->sc_crypto == G_ELI_CRYPTO_SW && g_eli_threads == 0) sched_bind(curthread, wr->w_number); thread_unlock(curthread); @@ -361,8 +361,7 @@ g_eli_worker(void *arg) mtx_unlock(&sc->sc_queue_mtx); kproc_exit(0); } - msleep(sc, &sc->sc_queue_mtx, PRIBIO | PDROP, - "geli:w", 0); + msleep(sc, &sc->sc_queue_mtx, PDROP, "geli:w", 0); continue; } mtx_unlock(&sc->sc_queue_mtx); ___ 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: r206666 - head/sbin/geom/misc
Author: pjd Date: Thu Apr 15 16:35:34 2010 New Revision: 20 URL: http://svn.freebsd.org/changeset/base/20 Log: Flush disk write cache after storing and clearing metadata. Modified: head/sbin/geom/misc/subr.c Modified: head/sbin/geom/misc/subr.c == --- head/sbin/geom/misc/subr.c Thu Apr 15 16:34:06 2010(r206665) +++ head/sbin/geom/misc/subr.c Thu Apr 15 16:35:34 2010(r20) @@ -236,6 +236,7 @@ g_metadata_store(const char *name, u_cha error = errno; goto out; } + (void)ioctl(fd, DIOCGFLUSH, NULL); out: if (sector != NULL) free(sector); @@ -293,6 +294,7 @@ g_metadata_clear(const char *name, const error = errno; goto out; } + (void)ioctl(fd, DIOCGFLUSH, NULL); out: if (sector != NULL) free(sector); ___ 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: r206667 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: pjd
Date: Thu Apr 15 16:40:54 2010
New Revision: 206667
URL: http://svn.freebsd.org/changeset/base/206667
Log:
Fix 3-way deadlock that can happen because of ZFS and vnode lock
order reversal.
thread0 (vfs_fhtovp) thread1 (vop_getattr) thread2 (zfs_recv)
- --
vn_lock
rrw_enter_read
rrw_enter_write (hangs)
rrw_enter_read (hangs)
vn_lock (hangs)
Submitted by: Attila Nagy
MFC after:3 days
Modified:
head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.c
head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c
Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.c
==
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.cThu Apr
15 16:35:34 2010(r20)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.cThu Apr
15 16:40:54 2010(r206667)
@@ -868,13 +868,15 @@ zfs_root(vfs_t *vfsp, int flags, vnode_t
ZFS_ENTER_NOERROR(zfsvfs);
error = zfs_zget(zfsvfs, zfsvfs->z_root, &rootzp);
+
+ ZFS_EXIT(zfsvfs);
+
if (error == 0) {
*vpp = ZTOV(rootzp);
error = vn_lock(*vpp, flags);
(*vpp)->v_vflag |= VV_ROOT;
}
- ZFS_EXIT(zfsvfs);
return (error);
}
@@ -1143,13 +1145,13 @@ zfs_vget(vfs_t *vfsp, ino_t ino, int fla
VN_RELE(ZTOV(zp));
err = EINVAL;
}
+ ZFS_EXIT(zfsvfs);
if (err != 0)
*vpp = NULL;
else {
*vpp = ZTOV(zp);
vn_lock(*vpp, flags);
}
- ZFS_EXIT(zfsvfs);
return (err);
}
@@ -1237,8 +1239,8 @@ zfs_fhtovp(vfs_t *vfsp, fid_t *fidp, vno
} else {
VN_HOLD(*vpp);
}
- vn_lock(*vpp, LK_EXCLUSIVE | LK_RETRY);
ZFS_EXIT(zfsvfs);
+ vn_lock(*vpp, LK_EXCLUSIVE | LK_RETRY);
return (0);
}
@@ -1259,10 +1261,11 @@ zfs_fhtovp(vfs_t *vfsp, fid_t *fidp, vno
return (EINVAL);
}
+ ZFS_EXIT(zfsvfs);
+
*vpp = ZTOV(zp);
vn_lock(*vpp, LK_EXCLUSIVE | LK_RETRY);
vnode_create_vobject(*vpp, zp->z_phys->zp_size, curthread);
- ZFS_EXIT(zfsvfs);
return (0);
}
Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c
==
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c Thu Apr
15 16:35:34 2010(r20)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c Thu Apr
15 16:40:54 2010(r206667)
@@ -1209,15 +1209,17 @@ zfs_lookup(vnode_t *dvp, char *nm, vnode
ltype = VOP_ISLOCKED(dvp);
VOP_UNLOCK(dvp, 0);
}
+ ZFS_EXIT(zfsvfs);
error = vn_lock(*vpp, cnp->cn_lkflags);
if (cnp->cn_flags & ISDOTDOT)
vn_lock(dvp, ltype | LK_RETRY);
if (error != 0) {
VN_RELE(*vpp);
*vpp = NULL;
- ZFS_EXIT(zfsvfs);
return (error);
}
+ } else {
+ ZFS_EXIT(zfsvfs);
}
#ifdef FREEBSD_NAMECACHE
@@ -1237,8 +1239,6 @@ zfs_lookup(vnode_t *dvp, char *nm, vnode
}
#endif
- ZFS_EXIT(zfsvfs);
-
return (error);
}
___
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: r206668 - head/etc/mtree
Author: pjd Date: Thu Apr 15 16:43:43 2010 New Revision: 206668 URL: http://svn.freebsd.org/changeset/base/206668 Log: Use spaces instead of tab for indent here. Modified: head/etc/mtree/BSD.include.dist Modified: head/etc/mtree/BSD.include.dist == --- head/etc/mtree/BSD.include.dist Thu Apr 15 16:40:54 2010 (r206667) +++ head/etc/mtree/BSD.include.dist Thu Apr 15 16:43:43 2010 (r206668) @@ -104,8 +104,8 @@ .. lmc .. - mfi - .. +mfi +.. mpt mpilib .. ___ 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: r206497 - in head: sbin/geom/class sbin/geom/class/sched sys/geom/sched sys/modules/geom sys/modules/geom/geom_sched sys/modules/geom/geom_sched/gs_sched sys/modules/geom/geom_sched/gs
On Thu, 15.04.2010 at 17:15:57 +0200, Luigi Rizzo wrote: > On Thu, Apr 15, 2010 at 04:38:23PM +0200, Attila Nagy wrote: > > On 04/12/2010 06:37 PM, Luigi Rizzo wrote: > > > Bring in geom_sched, support for scheduling disk I/O requests > > > in a device independent manner. Also include an example anticipatory > > > scheduler, gsched_rr, which gives very nice performance improvements > > > in presence of competing random access patterns. > > > > > > This is joint work with Fabio Checconi, developed last year > > > and presented at BSDCan 2009. You can find details in the > > > README file or at > > > > > Are there any plans to bring in something like Linux's ionice (or even > > better solutions)? Is it possible to do it easily at all (the user and > > GEOM are seemingly very far from each other)? > > I can easily[*] modify the I/O priority/weight of process or threads, > but you have to keep in mind that once a request is intercepted by > the filesystem code (e.g. for delayed writes, or avoid duplication > of pending reads), by the time it hits the disk it is not trivial > to tell who should be charged for it. > > This is an area where we still need to experiment a bit > (especially to find on which process issues which request, > say for metadata, delayed writes, shared reads). > > [*] the 'easy' way is put the information in an external hash table > and look it up using as a search key either of uid, gid, thread *>. I think we did something like this in the GSOC2005 > project on disk scheduling. A more intrusive way involves extending > the struct thread with an extra field to store the io priority. What would already be helpful, is an I/O time measurement of the overall system and have that displayed in, eg., vmstat like Linux and Solaris do. It gives a better measure of if a system is CPU bound or I/O bound or neither. Can the scheduler help here? Uli ___ 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: r206669 - head/sbin/hastd
Author: pjd Date: Thu Apr 15 17:04:08 2010 New Revision: 206669 URL: http://svn.freebsd.org/changeset/base/206669 Log: Increase ggate queue size to maximum value. HAST was not able to stand heavy random load. Reported by: Hiroyuki Yamagami MFC after:3 days Modified: head/sbin/hastd/primary.c Modified: head/sbin/hastd/primary.c == --- head/sbin/hastd/primary.c Thu Apr 15 16:43:43 2010(r206668) +++ head/sbin/hastd/primary.c Thu Apr 15 17:04:08 2010(r206669) @@ -682,7 +682,7 @@ init_ggate(struct hast_resource *res) ggiocreate.gctl_mediasize = res->hr_datasize; ggiocreate.gctl_sectorsize = res->hr_local_sectorsize; ggiocreate.gctl_flags = 0; - ggiocreate.gctl_maxcount = 128; + ggiocreate.gctl_maxcount = G_GATE_MAX_QUEUE_SIZE; ggiocreate.gctl_timeout = 0; ggiocreate.gctl_unit = G_GATE_NAME_GIVEN; snprintf(ggiocreate.gctl_name, sizeof(ggiocreate.gctl_name), "hast/%s", ___ 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: r206452 - head/sys/netinet
On 04/10/10 13:05, Bruce M Simpson wrote: Log: Fix a few issues related to the legacy 4.4 BSD multicast APIs. IPv4 addresses can and do change during normal operation. Testing by pfSense developers exposed an issue where OpenOSPFD was using the IPv4 address to leave the OSPF link-scope multicast groups on a dynamic OpenVPN tun interface, rather than using RFC 3678 with the interface index, which won't be raced when the interface's addresses change. I should point out that IP multicast is counter-intuitive in many areas. It seems that knowledge of how to work with it effectively is not that widespread. People who have a need to use it, often have very specific requirements. IP multicast group membership is always scoped to physical links [1]. The 4.4BSD API originally used the "primary IP address" to identify each link. Unfortunately this is not a persistent identifier, especially so in the use-case which had problems. In the updated API specified in RFC 3678, the primary key for multicast memberships, changes to the interface index. routed(8) in FreeBSD's SVN tree uses this new API; thanks to phk for reviewing and testing my change. In the case of OpenOSPFD+OpenVPN, the issue manifested itself in situations where this "primary IP address" changed. OpenVPN shims address assignment as part of its wire protocol, which is where the address change originates. This is purely a by-product of using an API which depends on using an IPv4 address as a key (IP_ADD_MEMBERSHIP/IP_DROP_MEMBERSHIP), when that key may change at any time. There's some time domain uncertainty with that change. PF_ROUTE is an out-of-band mechanism used by OpenOSPFD to learn about the address change, and it is dispatched separately from the change itself; there may be queueing and context switches involved. It also creates some ambiguity for in_multi instances, which the kernel may not be able to resolve, if that key has been blown away. Fortunately, such ambiguity is private to each socket. Furthermore, this makes IGMP somewhat unreliable, although mechanisms do exist in the protocol to work-around the end-station's identifier being lost in this way. The upstream IGMP querier (usually the mcast router) will hold state based on the end-station's "primary IP address", this will eventually time out due to unanswered queries. MLD side-steps this by always requiring an IPv6 link-scope address in control traffic. Using the more recent multicast RFCs, supported now by Windows, Linux, OpenSolaris, and now FreeBSD, can side-step these issues; it is the preferred API. P.S. HEAD was probably already resilient against this for joins. I plan to MFC a similar version of this change to 8 soon. [1] sendto() to an IP multicast address, without an interface specified, makes no sense -- unless the host is also an mcast forwarder. You can get away with default route resolution through the routing tables up to a point, but if your box is multihomed, all bets are off. ___ 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: r206671 - head/sys/kern
Author: kib
Date: Thu Apr 15 17:17:02 2010
New Revision: 206671
URL: http://svn.freebsd.org/changeset/base/206671
Log:
Fix typo.
MFC after:3 days
Modified:
head/sys/kern/vfs_cache.c
Modified: head/sys/kern/vfs_cache.c
==
--- head/sys/kern/vfs_cache.c Thu Apr 15 17:10:46 2010(r206670)
+++ head/sys/kern/vfs_cache.c Thu Apr 15 17:17:02 2010(r206671)
@@ -610,7 +610,7 @@ cache_enter(dvp, vp, cnp)
CTR3(KTR_VFS, "cache_enter(%p, %p, %s)", dvp, vp, cnp->cn_nameptr);
VNASSERT(vp == NULL || (vp->v_iflag & VI_DOOMED) == 0, vp,
- ("cahe_enter: Adding a doomed vnode"));
+ ("cache_enter: Adding a doomed vnode"));
if (!doingcache)
return;
___
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: r206672 - head/sys/dev/sge
Author: yongari
Date: Thu Apr 15 17:24:21 2010
New Revision: 206672
URL: http://svn.freebsd.org/changeset/base/206672
Log:
Fix include path.
Modified:
head/sys/dev/sge/if_sge.c
Modified: head/sys/dev/sge/if_sge.c
==
--- head/sys/dev/sge/if_sge.c Thu Apr 15 17:17:02 2010(r206671)
+++ head/sys/dev/sge/if_sge.c Thu Apr 15 17:24:21 2010(r206672)
@@ -81,7 +81,7 @@ __FBSDID("$FreeBSD$");
#include
#include
-#include "if_sgereg.h"
+#include
MODULE_DEPEND(sge, pci, 1, 1, 1);
MODULE_DEPEND(sge, ether, 1, 1, 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"
Re: svn commit: r206497 - in head: sbin/geom/class sbin/geom/class/sched sys/geom/sched sys/modules/geom sys/modules/geom/geom_sched sys/modules/geom/geom_sched/gs_sched sys/modules/geom/geom_sched/gs
In message <20100415164922.gd92...@acme.spoerlein.net>, Ulrich =?utf-8?B?U3DDtn JsZWlu?= writes: >What would already be helpful, is an I/O time measurement of the overall >system and have that displayed in, eg., vmstat like Linux and Solaris >do. It gives a better measure of if a system is CPU bound or I/O bound >or neither. gstat(8) ? -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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: r206675 - head/contrib/netcat
Author: delphij
Date: Thu Apr 15 19:15:05 2010
New Revision: 206675
URL: http://svn.freebsd.org/changeset/base/206675
Log:
Diff reduction with OpenBSD:
- Remove unused locally added variable;
- Deprecate -o: it's the default behavior on OpenBSD.
X-MFC:along with nc 4.7.
Modified:
head/contrib/netcat/nc.1
head/contrib/netcat/netcat.c
Modified: head/contrib/netcat/nc.1
==
--- head/contrib/netcat/nc.1Thu Apr 15 18:30:21 2010(r206674)
+++ head/contrib/netcat/nc.1Thu Apr 15 19:15:05 2010(r206675)
@@ -27,7 +27,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd February 23, 2010
+.Dd April 15, 2010
.Dt NC 1
.Os
.Sh NAME
@@ -36,7 +36,7 @@
.Sh SYNOPSIS
.Nm nc
.Bk -words
-.Op Fl 46DdEhklnorStUuvz
+.Op Fl 46DdEhklnrStUuvz
.Op Fl e Ar IPsec_policy
.Op Fl I Ar length
.Op Fl i Ar interval
@@ -159,15 +159,6 @@ socket option.
.It Fl O Ar length
Specifies the size of the TCP send buffer.
When
-.It Fl o
-.Dq Once-only mode .
-By default,
-.Nm
-does not terminate on EOF condition on input,
-but continues until the network side has been closed down.
-Specifying
-.Fl o
-will make it terminate on EOF as well.
.It Fl P Ar proxy_username
Specifies a username to present to a proxy server that requires authentication.
If no username is specified then authentication will not be attempted.
Modified: head/contrib/netcat/netcat.c
==
--- head/contrib/netcat/netcat.cThu Apr 15 18:30:21 2010
(r206674)
+++ head/contrib/netcat/netcat.cThu Apr 15 19:15:05 2010
(r206675)
@@ -72,14 +72,12 @@
#define PORT_MAX_LEN 6
/* Command Line Options */
-intEflag; /* Use IPsec ESP */
intdflag; /* detached, no stdin */
unsigned int iflag;/* Interval Flag */
intjflag; /* use jumbo frames if we can */
intkflag; /* More than one connect */
intlflag; /* Bind to local port */
intnflag; /* Don't do name look up */
-intoflag; /* Once only: stop on EOF */
intFreeBSD_Oflag; /* Do not use TCP options */
char *Pflag; /* Proxy username */
char *pflag; /* Localport flag */
@@ -151,7 +149,7 @@ main(int argc, char *argv[])
sv = NULL;
while ((ch = getopt_long(argc, argv,
- "46DdEe:hI:i:jklnO:oP:p:rSs:tT:UuV:vw:X:x:z",
+ "46DdEe:hI:i:jklnoO:P:p:rSs:tT:UuV:vw:X:x:z",
longopts, NULL)) != -1) {
switch (ch) {
case '4':
@@ -214,7 +212,7 @@ main(int argc, char *argv[])
nflag = 1;
break;
case 'o':
- oflag = 1;
+ fprintf(stderr, "option -o is deprecated.\n");
break;
case 'P':
Pflag = optarg;
@@ -282,8 +280,6 @@ main(int argc, char *argv[])
case 'T':
Tflag = parse_iptos(optarg);
break;
- case 0:
- break;
default:
usage(1);
}
@@ -729,10 +725,9 @@ readwrite(int nfd)
}
if (!dflag && pfd[1].revents & POLLIN) {
- if ((n = read(wfd, buf, plen)) < 0 ||
- (oflag && n == 0)) {
+ if ((n = read(wfd, buf, plen)) < 0)
return;
- } else if (n == 0) {
+ else if (n == 0) {
shutdown(nfd, SHUT_WR);
pfd[1].fd = -1;
pfd[1].events = 0;
@@ -945,7 +940,6 @@ help(void)
\t-nSuppress name/port resolutions\n\
\t--no-tcpopt Disable TCP options\n\
\t-O length TCP send buffer length\n\
- \t-oTerminate on EOF on input\n\
\t-P proxyuser\tUsername for proxy authentication\n\
\t-p port\t Specify local port for remote connects\n\
\t-rRandomize remote ports\n\
@@ -995,9 +989,9 @@ usage(int ret)
{
fprintf(stderr,
#ifdef IPSEC
- "usage: nc [-46DdEhklnorStUuvz] [-e policy] [-I length] [-i
interval] [-O length]\n"
+ "usage: nc [-46DdEhklnrStUuvz] [-e policy] [-I length] [-i
interval] [-O length]\n"
#else
- "usage: nc [-46DdhklnorStUuvz] [-I length] [-i interval] [-O
length]\n"
+ "usage: nc [-46DdhklnrStUuvz] [-I length] [-i interval] [-O
length]\n"
#endif
"\t
svn commit: r206684 - head/sys/dev/hwpmc
Author: fabient
Date: Thu Apr 15 19:45:03 2010
New Revision: 206684
URL: http://svn.freebsd.org/changeset/base/206684
Log:
- Fix a typo OFFCORE_REQUESTS.ANY.RFO is B0H10H and not 80H10H.
- Enable missing PARTIAL_ADDRESS_ALIAS for Core i7.
MFC after: 3 days
Modified:
head/sys/dev/hwpmc/hwpmc_core.c
head/sys/dev/hwpmc/pmc_events.h
Modified: head/sys/dev/hwpmc/hwpmc_core.c
==
--- head/sys/dev/hwpmc/hwpmc_core.c Thu Apr 15 19:31:21 2010
(r206683)
+++ head/sys/dev/hwpmc/hwpmc_core.c Thu Apr 15 19:45:03 2010
(r206684)
@@ -603,7 +603,7 @@ static struct iap_event_descr iap_events
IAPDESCR(06H_0FH, 0x06, 0x0F, IAP_F_FM | IAP_F_I7O),
IAPDESCR(07H_00H, 0x07, 0x00, IAP_F_FM | IAP_F_CC | IAP_F_CC2),
-IAPDESCR(07H_01H, 0x07, 0x01, IAP_F_FM | IAP_F_ALLCPUSCORE2 | IAP_F_WM),
+IAPDESCR(07H_01H, 0x07, 0x01, IAP_F_FM | IAP_F_ALLCPUSCORE2 | IAP_F_I7 |
IAP_F_WM),
IAPDESCR(07H_02H, 0x07, 0x02, IAP_F_FM | IAP_F_ALLCPUSCORE2),
IAPDESCR(07H_03H, 0x07, 0x03, IAP_F_FM | IAP_F_ALLCPUSCORE2),
IAPDESCR(07H_06H, 0x07, 0x06, IAP_F_FM | IAP_F_CA),
@@ -1053,6 +1053,7 @@ static struct iap_event_descr iap_events
IAPDESCR(B0H_02H, 0xB0, 0x02, IAP_F_FM | IAP_F_WM | IAP_F_I7O),
IAPDESCR(B0H_04H, 0xB0, 0x04, IAP_F_FM | IAP_F_WM | IAP_F_I7O),
IAPDESCR(B0H_08H, 0xB0, 0x08, IAP_F_FM | IAP_F_WM | IAP_F_I7O),
+IAPDESCR(B0H_10H, 0xB0, 0x10, IAP_F_FM | IAP_F_WM | IAP_F_I7O),
IAPDESCR(B0H_20H, 0xB0, 0x20, IAP_F_FM | IAP_F_I7O),
IAPDESCR(B0H_40H, 0xB0, 0x40, IAP_F_FM | IAP_F_I7 | IAP_F_WM),
IAPDESCR(B0H_80H, 0xB0, 0x80, IAP_F_FM | IAP_F_CA | IAP_F_WM | IAP_F_I7O),
Modified: head/sys/dev/hwpmc/pmc_events.h
==
--- head/sys/dev/hwpmc/pmc_events.h Thu Apr 15 19:31:21 2010
(r206683)
+++ head/sys/dev/hwpmc/pmc_events.h Thu Apr 15 19:45:03 2010
(r206684)
@@ -2207,7 +2207,7 @@ __PMC_EV_ALIAS("OFFCORE_REQUESTS.DEMAND.
__PMC_EV_ALIAS("OFFCORE_REQUESTS.DEMAND.READ_CODE", IAP_EVENT_B0H_02H) \
__PMC_EV_ALIAS("OFFCORE_REQUESTS.DEMAND.RFO", IAP_EVENT_B0H_04H) \
__PMC_EV_ALIAS("OFFCORE_REQUESTS.ANY.READ", IAP_EVENT_B0H_08H) \
-__PMC_EV_ALIAS("OFFCORE_REQUESTS.ANY.RFO", IAP_EVENT_80H_10H) \
+__PMC_EV_ALIAS("OFFCORE_REQUESTS.ANY.RFO", IAP_EVENT_B0H_10H) \
__PMC_EV_ALIAS("OFFCORE_REQUESTS.L1D_WRITEBACK", IAP_EVENT_B0H_40H)\
__PMC_EV_ALIAS("OFFCORE_REQUESTS.ANY", IAP_EVENT_B0H_80H) \
__PMC_EV_ALIAS("UOPS_EXECUTED.PORT0", IAP_EVENT_B1H_01H) \
___
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: r206686 - head/etc
Author: dougb
Date: Thu Apr 15 21:18:24 2010
New Revision: 206686
URL: http://svn.freebsd.org/changeset/base/206686
Log:
If a service is running, make 'stop' work even if ${name}_enable
is not set.
PR: conf/130414
Submitted by: Dominic Fandrey
Reviewed by: freebsd-rc@
Modified:
head/etc/rc.subr
Modified: head/etc/rc.subr
==
--- head/etc/rc.subrThu Apr 15 20:48:57 2010(r206685)
+++ head/etc/rc.subrThu Apr 15 21:18:24 2010(r206686)
@@ -646,12 +646,12 @@ run_rc_command()
if [ "$_elem" != "$rc_arg" ]; then
continue
fi
- # if ${rcvar} is set, and $1 is not
- # "rcvar", then run
+ # if ${rcvar} is set, $1 is not "rcvar"
+ # and ${rc_pid} is not set, then run
# checkyesno ${rcvar}
# and return if that failed
#
- if [ -n "${rcvar}" -a "$rc_arg" != "rcvar" ]; then
+ if [ -n "${rcvar}" -a "$rc_arg" != "rcvar" -a -z "${rc_pid}" ];
then
if ! checkyesno ${rcvar}; then
if [ -n "${rc_quiet}" ]; then
return 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: r205954 - head/include
On Wed, Mar 31, 2010, Juli Mallett wrote: > Be like stdlib.h and bring in a wchar_t definition and use it to prototype > wcstoimax and wcstoumax, rather than spelling it __wchar_t. This is > necessary > to use these functions in C++ where wchar_t is different to __wchar_t and is > a built-in type. I don't think this change is correct for C mode. POSIX says wchar_t is *supposed* to be defined by stdlib.h and wchar.h, but its definition is namespace pollution elsewhere. > It may be better to use __wchar_t here and to simply define __wchar_t as > being > wchar_t in C++ mode rather than to bring in wchar_t That may be better. ___ 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: r206687 - head/usr.bin/indent
Author: avg
Date: Thu Apr 15 21:41:07 2010
New Revision: 206687
URL: http://svn.freebsd.org/changeset/base/206687
Log:
indent(1): don't treat bare '_t' as a type name when -ta is used
It seems that identifier "_t" is sometimes used as a variable name,
even in our tree. Not that I endorse that, but still it's better
to require at least one character before _t suffix to consider
an identifier to be a type name.
Reported by: Alex Vasylenko
MFC after:1 week
Modified:
head/usr.bin/indent/lexi.c
Modified: head/usr.bin/indent/lexi.c
==
--- head/usr.bin/indent/lexi.c Thu Apr 15 21:18:24 2010(r206686)
+++ head/usr.bin/indent/lexi.c Thu Apr 15 21:41:07 2010(r206687)
@@ -251,9 +251,10 @@ lexi(void)
if (auto_typedefs) {
const char *q = s_token;
+ size_t q_len = strlen(q);
/* Check if we have an "_t" in the end */
- if (q[0] && q[1] &&
- (strcmp(q + strlen(q) - 2, "_t") == 0)) {
+ if (q_len > 2 &&
+ (strcmp(q + q_len - 2, "_t") == 0)) {
ps.its_a_keyword = true;
ps.last_u_d = true;
goto found_auto_typedef;
___
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: r206688 - in head/sys/fs: nfs nfsclient
Author: rmacklem
Date: Thu Apr 15 22:57:30 2010
New Revision: 206688
URL: http://svn.freebsd.org/changeset/base/206688
Log:
The experimental NFS client was not filling in recovery credentials
for opens done locally in the client when a delegation for the file
was held. This could cause the client to crash in crsetgroups() when
recovering from a server crash/reboot. This patch fills in the
recovery credentials for this case, in order to avoid the client crash.
Also, add KASSERT()s to the credential copy functions, to catch any
other cases where the credentials aren't filled in correctly.
MFC after:1 week
Modified:
head/sys/fs/nfs/nfs_commonport.c
head/sys/fs/nfs/nfsclstate.h
head/sys/fs/nfsclient/nfs_clport.c
head/sys/fs/nfsclient/nfs_clrpcops.c
head/sys/fs/nfsclient/nfs_clstate.c
Modified: head/sys/fs/nfs/nfs_commonport.c
==
--- head/sys/fs/nfs/nfs_commonport.cThu Apr 15 21:41:07 2010
(r206687)
+++ head/sys/fs/nfs/nfs_commonport.cThu Apr 15 22:57:30 2010
(r206688)
@@ -225,6 +225,8 @@ void
newnfs_copycred(struct nfscred *nfscr, struct ucred *cr)
{
+ KASSERT(nfscr->nfsc_ngroups >= 0,
+ ("newnfs_copycred: negative nfsc_ngroups"));
cr->cr_uid = nfscr->nfsc_uid;
crsetgroups(cr, nfscr->nfsc_ngroups, nfscr->nfsc_groups);
}
Modified: head/sys/fs/nfs/nfsclstate.h
==
--- head/sys/fs/nfs/nfsclstate.hThu Apr 15 21:41:07 2010
(r206687)
+++ head/sys/fs/nfs/nfsclstate.hThu Apr 15 22:57:30 2010
(r206688)
@@ -140,6 +140,7 @@ struct nfsclopen {
#defineNFSCLOPEN_OK0
#defineNFSCLOPEN_DOOPEN1
#defineNFSCLOPEN_DOOPENDOWNGRADE 2
+#defineNFSCLOPEN_SETCRED 3
struct nfscllockowner {
LIST_ENTRY(nfscllockowner) nfsl_list;
Modified: head/sys/fs/nfsclient/nfs_clport.c
==
--- head/sys/fs/nfsclient/nfs_clport.c Thu Apr 15 21:41:07 2010
(r206687)
+++ head/sys/fs/nfsclient/nfs_clport.c Thu Apr 15 22:57:30 2010
(r206688)
@@ -978,6 +978,8 @@ newnfs_copyincred(struct ucred *cr, stru
{
int i;
+ KASSERT(cr->cr_ngroups >= 0,
+ ("newnfs_copyincred: negative cr_ngroups"));
nfscr->nfsc_uid = cr->cr_uid;
nfscr->nfsc_ngroups = MIN(cr->cr_ngroups, NFS_MAXGRPS + 1);
for (i = 0; i < nfscr->nfsc_ngroups; i++)
Modified: head/sys/fs/nfsclient/nfs_clrpcops.c
==
--- head/sys/fs/nfsclient/nfs_clrpcops.cThu Apr 15 21:41:07 2010
(r206687)
+++ head/sys/fs/nfsclient/nfs_clrpcops.cThu Apr 15 22:57:30 2010
(r206688)
@@ -278,7 +278,13 @@ else printf(" fhl=0\n");
error = EIO;
}
newnfs_copyincred(cred, &op->nfso_cred);
- }
+ } else if (ret == NFSCLOPEN_SETCRED)
+ /*
+* This is a new local open on a delegation. It needs
+* to have credentials so that an open can be done
+* against the server during recovery.
+*/
+ newnfs_copyincred(cred, &op->nfso_cred);
/*
* nfso_opencnt is the count of how many VOP_OPEN()s have
Modified: head/sys/fs/nfsclient/nfs_clstate.c
==
--- head/sys/fs/nfsclient/nfs_clstate.c Thu Apr 15 21:41:07 2010
(r206687)
+++ head/sys/fs/nfsclient/nfs_clstate.c Thu Apr 15 22:57:30 2010
(r206688)
@@ -274,8 +274,13 @@ nfscl_open(vnode_t vp, u_int8_t *nfhp, i
*owpp = owp;
if (opp != NULL)
*opp = op;
- if (retp != NULL)
- *retp = NFSCLOPEN_OK;
+ if (retp != NULL) {
+ if (nfhp != NULL && dp != NULL && nop == NULL)
+ /* new local open on delegation */
+ *retp = NFSCLOPEN_SETCRED;
+ else
+ *retp = NFSCLOPEN_OK;
+ }
/*
* Now, check the mode on the open and return the appropriate
___
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: r206689 - head/contrib/netcat
Author: delphij
Date: Thu Apr 15 23:21:24 2010
New Revision: 206689
URL: http://svn.freebsd.org/changeset/base/206689
Log:
Utilize IP_BINDANY which provided the same semantics of OpenBSD's
SO_BINDANY.
Modified:
head/contrib/netcat/netcat.c
Modified: head/contrib/netcat/netcat.c
==
--- head/contrib/netcat/netcat.cThu Apr 15 22:57:30 2010
(r206688)
+++ head/contrib/netcat/netcat.cThu Apr 15 23:21:24 2010
(r206689)
@@ -570,10 +570,8 @@ remote_connect(const char *host, const c
if (sflag || pflag) {
struct addrinfo ahints, *ares;
-#ifdef SO_BINDANY
- /* try SO_BINDANY, but don't insist */
- setsockopt(s, SOL_SOCKET, SO_BINDANY, &on, sizeof(on));
-#endif
+ /* try IP_BINDANY, but don't insist */
+ setsockopt(s, IPPROTO_IP, IP_BINDANY, &on, sizeof(on));
memset(&ahints, 0, sizeof(struct addrinfo));
ahints.ai_family = res0->ai_family;
ahints.ai_socktype = uflag ? SOCK_DGRAM : SOCK_STREAM;
___
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: r206690 - head/sys/fs/nfsclient
Author: rmacklem
Date: Thu Apr 15 23:56:05 2010
New Revision: 206690
URL: http://svn.freebsd.org/changeset/base/206690
Log:
Add mutex lock calls to 2 cases in the experimental NFS client's
renew thread where they were missing.
MFC after:1 week
Modified:
head/sys/fs/nfsclient/nfs_clstate.c
Modified: head/sys/fs/nfsclient/nfs_clstate.c
==
--- head/sys/fs/nfsclient/nfs_clstate.c Thu Apr 15 23:21:24 2010
(r206689)
+++ head/sys/fs/nfsclient/nfs_clstate.c Thu Apr 15 23:56:05 2010
(r206690)
@@ -2318,7 +2318,9 @@ nfscl_renewthread(struct nfsclclient *cl
int error, cbpathdown, islept, igotlock, ret, clearok;
cred = newnfs_getcred();
+ NFSLOCKCLSTATE();
clp->nfsc_flags |= NFSCLFLAGS_HASTHREAD;
+ NFSUNLOCKCLSTATE();
for(;;) {
newnfs_setroot(cred);
cbpathdown = 0;
@@ -2331,9 +2333,11 @@ nfscl_renewthread(struct nfsclclient *cl
error = nfsrpc_renew(clp, cred, p);
if (error == NFSERR_CBPATHDOWN)
cbpathdown = 1;
- else if (error == NFSERR_STALECLIENTID)
+ else if (error == NFSERR_STALECLIENTID) {
+ NFSLOCKCLSTATE();
clp->nfsc_flags |= NFSCLFLAGS_RECOVER;
- else if (error == NFSERR_EXPIRED)
+ NFSUNLOCKCLSTATE();
+ } else if (error == NFSERR_EXPIRED)
(void) nfscl_hasexpired(clp, clidrev, p);
}
___
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: r206695 - in head: . sys/mips/include sys/mips/mips
Author: jmallett
Date: Fri Apr 16 02:56:24 2010
New Revision: 206695
URL: http://svn.freebsd.org/changeset/base/206695
Log:
Remove some unused header files.
Deleted:
head/sys/mips/include/archtype.h
head/sys/mips/include/defs.h
head/sys/mips/include/queue.h
head/sys/mips/include/rm7000.h
head/sys/mips/include/segments.h
Modified:
head/ObsoleteFiles.inc
head/sys/mips/mips/trap.c
Modified: head/ObsoleteFiles.inc
==
--- head/ObsoleteFiles.inc Fri Apr 16 02:16:58 2010(r206694)
+++ head/ObsoleteFiles.inc Fri Apr 16 02:56:24 2010(r206695)
@@ -14,6 +14,14 @@
# The file is partitioned: OLD_FILES first, then OLD_LIBS and OLD_DIRS last.
#
+# 20100415: [mips] removed unused headers
+.if ${TARGET_ARCH} == "mips"
+OLD_FILES+=usr/include/machine/archtype.h
+OLD_FILES+=usr/include/machine/segments.h
+OLD_FILES+=usr/include/machine/rm7000.h
+OLD_FILES+=usr/include/machine/defs.h
+OLD_FILES+=usr/include/machine/queue.h
+.endif
# 20100326: [ia64] removed
.if ${TARGET_ARCH} == "ia64"
OLD_FILES+=usr/include/machine/nexusvar.h
Modified: head/sys/mips/mips/trap.c
==
--- head/sys/mips/mips/trap.c Fri Apr 16 02:16:58 2010(r206694)
+++ head/sys/mips/mips/trap.c Fri Apr 16 02:56:24 2010(r206695)
@@ -83,8 +83,6 @@ __FBSDID("$FreeBSD$");
#include
#include
#include
-#include
-#include
#include
#ifdef DDB
___
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: r206696 - head/sbin/hastd
Author: pjd
Date: Fri Apr 16 06:47:29 2010
New Revision: 206696
URL: http://svn.freebsd.org/changeset/base/206696
Log:
Fix control socket leak when worker process exits.
Submitted by: Mikolaj Golub
MFC after:3 days
Modified:
head/sbin/hastd/hastd.c
Modified: head/sbin/hastd/hastd.c
==
--- head/sbin/hastd/hastd.c Fri Apr 16 02:56:24 2010(r206695)
+++ head/sbin/hastd/hastd.c Fri Apr 16 06:47:29 2010(r206696)
@@ -137,6 +137,7 @@ child_exit(void)
pjdlog_error("Worker process failed (pid=%u,
status=%d).",
(unsigned int)pid, WEXITSTATUS(status));
}
+ proto_close(res->hr_ctrl);
res->hr_workerpid = 0;
if (res->hr_role == HAST_ROLE_PRIMARY) {
sleep(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: r206697 - head/sbin/hastd
Author: pjd
Date: Fri Apr 16 06:49:12 2010
New Revision: 206697
URL: http://svn.freebsd.org/changeset/base/206697
Log:
Fix log size calculation which caused message truncation.
Submitted by: Mikolaj Golub
MFC after:3 days
Modified:
head/sbin/hastd/pjdlog.c
Modified: head/sbin/hastd/pjdlog.c
==
--- head/sbin/hastd/pjdlog.cFri Apr 16 06:47:29 2010(r206696)
+++ head/sbin/hastd/pjdlog.cFri Apr 16 06:49:12 2010(r206697)
@@ -228,7 +228,7 @@ pjdlogv_common(int loglevel, int debugle
len = snprintf(log, sizeof(log), "%s", pjdlog_prefix);
if ((size_t)len < sizeof(log))
- len = vsnprintf(log + len, sizeof(log) - len, fmt, ap);
+ len += vsnprintf(log + len, sizeof(log) - len, fmt, ap);
if (error != -1 && (size_t)len < sizeof(log)) {
(void)snprintf(log + len, sizeof(log) - len, ": %s.",
strerror(error));
___
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"
