Re: svn commit: r254585 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
on 28/08/2013 04:02 Xin Li said the following: I'm not sure that this is right. Now we have: tdzp = VTOZ(tdvp); ZFS_VERIFY_ZP(tdzp); zfsvfs = tdzp-z_zfsvfs; ZFS_ENTER(zfsvfs);// tdzp's z_zfsvfs entered zilog = zfsvfs-z_log; sdzp = VTOZ(sdvp); ZFS_VERIFY_ZP(sdzp); // (*) Note that in the (*) step, when sdzp is invalid and sdzp have different z_zfsvfs than tdzp (for instance when the node is in the snapshot directory; the code later would test this), we could end up with ZFS_EXIT()'ing the wrong z_zfsvfs. The v_vfsp / v_mount check should be prior to any of this and should ensure that all the vnodes have the same z_zfsvfs. -- Andriy Gapon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254982 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
on 28/08/2013 03:39 Xin LI said the following: @@ -6250,8 +6250,11 @@ zfs_freebsd_rename(ap) ASSERT(ap-a_fcnp-cn_flags (SAVENAME|SAVESTART)); ASSERT(ap-a_tcnp-cn_flags (SAVENAME|SAVESTART)); - error = zfs_rename(fdvp, ap-a_fcnp-cn_nameptr, tdvp, - ap-a_tcnp-cn_nameptr, ap-a_fcnp-cn_cred, NULL, 0); + if (fdvp-v_mount == tdvp-v_mount) + error = zfs_rename(fdvp, ap-a_fcnp-cn_nameptr, tdvp, + ap-a_tcnp-cn_nameptr, ap-a_fcnp-cn_cred, NULL, 0); + else + error = EXDEV; if (tdvp == tvp) VN_RELE(tdvp); So, I am still not sure if that is important or not, but this change still misses (tvp (fvp-v_mount != tvp-v_mount)) check as found in ufs_rename. -- Andriy Gapon ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254985 - head/sys/modules/ip6_mroute_mod
Author: uqs Date: Wed Aug 28 07:48:44 2013 New Revision: 254985 URL: http://svnweb.freebsd.org/changeset/base/254985 Log: Fix 'make depend' Modified: head/sys/modules/ip6_mroute_mod/Makefile Modified: head/sys/modules/ip6_mroute_mod/Makefile == --- head/sys/modules/ip6_mroute_mod/MakefileWed Aug 28 05:12:29 2013 (r254984) +++ head/sys/modules/ip6_mroute_mod/MakefileWed Aug 28 07:48:44 2013 (r254985) @@ -7,7 +7,7 @@ KMOD= ip6_mroute SRCS= ip6_mroute.c -SRCS+= opt_inet6.h opt_mrouting.h +SRCS+= opt_inet6.h opt_kdtrace.h opt_mrouting.h .if !defined(KERNBUILDDIR) opt_inet6.h: ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254986 - head/sys/ufs/ufs
Author: ivoras Date: Wed Aug 28 10:06:20 2013 New Revision: 254986 URL: http://svnweb.freebsd.org/changeset/base/254986 Log: Take a very small step toward the Century of the Anchovy by increasing the time dirhash entries stay in memory before being considered for eviction to 1 minute. Modified: head/sys/ufs/ufs/ufs_dirhash.c Modified: head/sys/ufs/ufs/ufs_dirhash.c == --- head/sys/ufs/ufs/ufs_dirhash.c Wed Aug 28 07:48:44 2013 (r254985) +++ head/sys/ufs/ufs/ufs_dirhash.c Wed Aug 28 10:06:20 2013 (r254986) @@ -85,7 +85,7 @@ SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_d static int ufs_dirhashlowmemcount = 0; SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_lowmemcount, CTLFLAG_RD, ufs_dirhashlowmemcount, 0, number of times low memory hook called); -static int ufs_dirhashreclaimage = 5; +static int ufs_dirhashreclaimage = 60; SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_reclaimage, CTLFLAG_RW, ufs_dirhashreclaimage, 0, max time in seconds of hash inactivity before deletion in low VM events); ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254585 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
On Wed, Aug 28, 2013 at 3:02 AM, Xin Li delp...@delphij.net wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512
[snip]
On 08/27/13 01:59, Davide Italiano wrote:
I've written a patch that attempts to fix the second problem.
http://people.freebsd.org/~davide/review/zfsrename_recycle.diff The
culprit is that we're dereferencing sdvp without the vnode lock
held, so data-stability is not guaranteed. tdvp, instead, is locked
at the entry point of VOP_RENAME() (at least according to
vop_rename_pre()) so it can be safely used. As pointed out by
Andriy ZFS has some different mechanisms wrt other file systems
that allow the vnode to not be recycled when it's referenced
(ZFS_ENTER), so once we get zfsvfs_t * from tdzp we can call
ZFS_ENTER and dereference sdvp in a safe manner.
I'm not sure that this is right. Now we have:
tdzp = VTOZ(tdvp);
ZFS_VERIFY_ZP(tdzp);
zfsvfs = tdzp-z_zfsvfs;
ZFS_ENTER(zfsvfs); // tdzp's z_zfsvfs entered
zilog = zfsvfs-z_log;
sdzp = VTOZ(sdvp);
ZFS_VERIFY_ZP(sdzp);// (*)
Well, if your concern is that the running thread can exit from a
different context than the one it entered, maybe partly inlining
ZFS_VERIFY_ZP() might workaround the problem.
if (sdzp-z_sa_hdl == NULL) {
ZFS_EXIT(zfsvfs); /* this is the right context */
return (EIO);
}
Does this make sense for you? If not, can you propose an alternative?
I'll be away for a couple of days but I will be happy to discuss this
further when I'll come back.
Note that in the (*) step, when sdzp is invalid and sdzp have
different z_zfsvfs than tdzp (for instance when the node is in the
snapshot directory; the code later would test this), we could end up
with ZFS_EXIT()'ing the wrong z_zfsvfs.
Cheers,
- --
Xin LI delp...@delphij.nethttps://www.delphij.net/
FreeBSD - The Power to Serve! Live free or die
-BEGIN PGP SIGNATURE-
Version: GnuPG v2.0.21 (FreeBSD)
iQEcBAEBCgAGBQJSHUxBAAoJEG80Jeu8UPuzkzEIAKXlfGuTodrPcPkDkxBhhaOj
QoxoT06jFwMTplysICuCpslQNyyG2Jxq654u9nMh6q5dww370eTtm2FJ0n2QTxk4
JeWLpZVUu7VHWNXYVJQqrmATaMFHO4wVf5AYpSHn+1iCWo0kQn1MPxPw/oSUt0yw
1628jhWTs8n+rxQaYrYN6ewXYeylMwC50ZB0kE/gQpQZ+cKAGmrM/8tur24SQBEo
WwrHakrv1DGJ8rv3Q53FB2iUZ+zEAZs6MJ28w32lV3vOI20jfQbEK+RR7i7HvxdV
c/7M96wCd0X4iEqZb87VVfJFSRdQsXy/35scXhhh/BSviT7rervS8XxPGhfpXOs=
=MzwT
-END PGP SIGNATURE-
Thanks,
--
Davide
There are no solved problems; there are only problems that are more
or less solved -- Henri Poincare
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On Wed, Aug 28, 2013 at 12:06 PM, Ivan Voras ivo...@freebsd.org wrote: Author: ivoras Date: Wed Aug 28 10:06:20 2013 New Revision: 254986 URL: http://svnweb.freebsd.org/changeset/base/254986 Log: Take a very small step toward the Century of the Anchovy by increasing the time dirhash entries stay in memory before being considered for eviction to 1 minute. Modified: head/sys/ufs/ufs/ufs_dirhash.c Modified: head/sys/ufs/ufs/ufs_dirhash.c == --- head/sys/ufs/ufs/ufs_dirhash.c Wed Aug 28 07:48:44 2013 (r254985) +++ head/sys/ufs/ufs/ufs_dirhash.c Wed Aug 28 10:06:20 2013 (r254986) @@ -85,7 +85,7 @@ SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_d static int ufs_dirhashlowmemcount = 0; SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_lowmemcount, CTLFLAG_RD, ufs_dirhashlowmemcount, 0, number of times low memory hook called); -static int ufs_dirhashreclaimage = 5; +static int ufs_dirhashreclaimage = 60; SYSCTL_INT(_vfs_ufs, OID_AUTO, dirhash_reclaimage, CTLFLAG_RW, ufs_dirhashreclaimage, 0, max time in seconds of hash inactivity before deletion in low VM events); Hi, do you have any evidence that this change impacts positively (or negatively) performances for some workloads? If yes, can you share? Also, why did you choose the '60' value (rather than something else)? I don't see any 'Reviewed by:' line in your commit message neither I remember a public discussion on -current or -arch or -fs about this. OTOH I think such changes deserve a wider discussion. Thanks, -- Davide There are no solved problems; there are only problems that are more or less solved -- Henri Poincare ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On 28 August 2013 12:25, Davide Italiano dav...@freebsd.org wrote: do you have any evidence that this change impacts positively (or negatively) performances for some workloads? If yes, can you share? Yes, observation of my own servers. Without this, dirhash is basically useless since in certain situations everything gets evicted after 5 seconds and it never grows to its full potential. Also, why did you choose the '60' value (rather than something else)? Personal experience. I don't see any 'Reviewed by:' line in your commit message neither I remember a public discussion on -current or -arch or -fs about this. OTOH I think such changes deserve a wider discussion. See discussion in @stable. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On Wed, Aug 28, 2013 at 12:31 PM, Ivan Voras ivo...@freebsd.org wrote: On 28 August 2013 12:25, Davide Italiano dav...@freebsd.org wrote: do you have any evidence that this change impacts positively (or negatively) performances for some workloads? If yes, can you share? Yes, observation of my own servers. Without this, dirhash is basically useless since in certain situations everything gets evicted after 5 seconds and it never grows to its full potential. Oh, well. Your servers doesn't necessary reflect FreeBSD general user so you can just maintain this change local. I've always thought FreeBSD policy was that of leaving reasonable defaults for the general case and giving people possibility to tune them according to their needs. Also 'in certain situations' makes relatively little sense to me unless you provide a testcase that mimics such situations or explain to people how to reproduce it. Also, why did you choose the '60' value (rather than something else)? Personal experience. Do you realize that this is a driven by commit change, right? And there's no technical motivatiion about this or experimental data that confirms you've chossen the right value? I don't see any 'Reviewed by:' line in your commit message neither I remember a public discussion on -current or -arch or -fs about this. OTOH I think such changes deserve a wider discussion. See discussion in @stable. I see from the archives that you didn't get any feedback by anyone working in the VM layer or by some UFS maintainer. I don't even comment about discussing -CURRENT changes in stable@. Thanks, -- Davide There are no solved problems; there are only problems that are more or less solved -- Henri Poincare ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On 28 August 2013 12:44, Davide Italiano dav...@freebsd.org wrote: Do you realize that this is a driven by commit change, right? And there's no technical motivatiion about this or experimental data that confirms you've chossen the right value? I see from the archives that you didn't get any feedback by anyone working in the VM layer or by some UFS maintainer. I don't even comment about discussing -CURRENT changes in stable@. You have a point that this was a judgment call, but note that the last change to dirhash tuning (maxmem) was also done by me. Since this is -CURRENT, I will leave this commit in unless it proves detrimental in the usual testing which goes on. If you still feel strongly about this, contact me in private and we'll investigate its effects in more detail. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On Wed, Aug 28, 2013 at 12:53 PM, Ivan Voras ivo...@freebsd.org wrote: On 28 August 2013 12:44, Davide Italiano dav...@freebsd.org wrote: Do you realize that this is a driven by commit change, right? And there's no technical motivatiion about this or experimental data that confirms you've chossen the right value? I see from the archives that you didn't get any feedback by anyone working in the VM layer or by some UFS maintainer. I don't even comment about discussing -CURRENT changes in stable@. You have a point that this was a judgment call, but note that the last change to dirhash tuning (maxmem) was also done by me. Since this is -CURRENT, I will leave this commit in unless it proves detrimental in the usual testing which goes on. If 10.0 would be one or two years away, I would be OK with that. But we're already in code slush and in 10 days -CURRENT will become -STABLE'ish so I'm worried this change won't have the right exposure. I would be more happy to see it postponed, maybe after proper discussion and testing. I'm not against the change as-it-is, my point is that the project reached a point of maturity where 'judgment calls' doesn't work very well. It's more a matter of how things are done rather than what things are done. If you still feel strongly about this, contact me in private and we'll investigate its effects in more detail. I prefer to see a public discussion about this. Maybe you should come up with some evidence that your change is helpful, even after this commit. -- Davide There are no solved problems; there are only problems that are more or less solved -- Henri Poincare ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254986 - head/sys/ufs/ufs
On 28 August 2013 13:14, Davide Italiano dav...@freebsd.org wrote: I prefer to see a public discussion about this. Ok, Ok, I'll go start a retrograde disucussion on @filesystems :) Maybe you should come up with some evidence that your change is helpful, even after this commit. I'll try and dig something up. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254987 - head/sys/dev/ahci
Author: gavin
Date: Wed Aug 28 14:29:33 2013
New Revision: 254987
URL: http://svnweb.freebsd.org/changeset/base/254987
Log:
Support the PCI-Express SSD in the new MacBook Air (model A1465)
Submitted by: Johannes Lundberg johannes brilliantservice.co.jp
MFC after:3 days
Modified:
head/sys/dev/ahci/ahci.c
Modified: head/sys/dev/ahci/ahci.c
==
--- head/sys/dev/ahci/ahci.cWed Aug 28 10:06:20 2013(r254986)
+++ head/sys/dev/ahci/ahci.cWed Aug 28 14:29:33 2013(r254987)
@@ -229,6 +229,7 @@ static struct {
{0x91301b4b, 0x00, Marvell 88SE9130, AHCI_Q_NOBSYRES|AHCI_Q_ALTSIG},
{0x91721b4b, 0x00, Marvell 88SE9172, AHCI_Q_NOBSYRES},
{0x91821b4b, 0x00, Marvell 88SE9182, AHCI_Q_NOBSYRES},
+ {0x91831b4b, 0x00, Marvell 88SS9183, AHCI_Q_NOBSYRES},
{0x91a01b4b, 0x00, Marvell 88SE91Ax, AHCI_Q_NOBSYRES},
{0x92151b4b, 0x00, Marvell 88SE9215, AHCI_Q_NOBSYRES},
{0x92201b4b, 0x00, Marvell 88SE9220, AHCI_Q_NOBSYRES|AHCI_Q_ALTSIG},
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254988 - head/sys/dev/gpio
Author: loos
Date: Wed Aug 28 14:39:24 2013
New Revision: 254988
URL: http://svnweb.freebsd.org/changeset/base/254988
Log:
Properly free gpiobus ivars when gpiobus_parse_pins() fails and also on
gpiobus detachment.
Suggested by: imp
Approved by: adrian (mentor)
Modified:
head/sys/dev/gpio/gpiobus.c
Modified: head/sys/dev/gpio/gpiobus.c
==
--- head/sys/dev/gpio/gpiobus.c Wed Aug 28 14:29:33 2013(r254987)
+++ head/sys/dev/gpio/gpiobus.c Wed Aug 28 14:39:24 2013(r254988)
@@ -151,6 +151,7 @@ gpiobus_parse_pins(struct gpiobus_softc
if (i = sc-sc_npins) {
device_printf(child,
invalid pin %d, max: %d\n, i, sc-sc_npins - 1);
+ free(devi-pins, M_DEVBUF);
return (EINVAL);
}
@@ -161,6 +162,7 @@ gpiobus_parse_pins(struct gpiobus_softc
if (sc-sc_pins_mapped[i]) {
device_printf(child,
warning: pin %d is already mapped\n, i);
+ free(devi-pins, M_DEVBUF);
return (EINVAL);
}
sc-sc_pins_mapped[i] = 1;
@@ -218,9 +220,12 @@ gpiobus_attach(device_t dev)
static int
gpiobus_detach(device_t dev)
{
- struct gpiobus_softc *sc = GPIOBUS_SOFTC(dev);
- int err;
+ struct gpiobus_softc *sc;
+ struct gpiobus_ivar *devi;
+ device_t *devlist;
+ int i, err, ndevs;
+ sc = GPIOBUS_SOFTC(dev);
KASSERT(mtx_initialized(sc-sc_mtx),
(gpiobus mutex not initialized));
GPIOBUS_LOCK_DESTROY(sc);
@@ -228,8 +233,17 @@ gpiobus_detach(device_t dev)
if ((err = bus_generic_detach(dev)) != 0)
return (err);
- /* detach and delete all children */
- device_delete_children(dev);
+ if ((err = device_get_children(dev, devlist, ndevs)) != 0)
+ return (err);
+ for (i = 0; i ndevs; i++) {
+ device_delete_child(dev, devlist[i]);
+ devi = GPIOBUS_IVAR(devlist[i]);
+ if (devi-pins) {
+ free(devi-pins, M_DEVBUF);
+ devi-pins = NULL;
+ }
+ }
+ free(devlist, M_TEMP);
if (sc-sc_pins_mapped) {
free(sc-sc_pins_mapped, M_DEVBUF);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254989 - head/sys/mips/conf
Author: loos Date: Wed Aug 28 14:43:04 2013 New Revision: 254989 URL: http://svnweb.freebsd.org/changeset/base/254989 Log: Add the default hints to make the GPIO pins, rf led and reset switch work out of the box on RouterStation. PR: 177832 Submitted by: Petko Bordjukov (bordju...@gmail.com) Approved by: adrian (mentor) Modified: head/sys/mips/conf/ROUTERSTATION.hints Modified: head/sys/mips/conf/ROUTERSTATION.hints == --- head/sys/mips/conf/ROUTERSTATION.hints Wed Aug 28 14:39:24 2013 (r254988) +++ head/sys/mips/conf/ROUTERSTATION.hints Wed Aug 28 14:43:04 2013 (r254989) @@ -23,6 +23,24 @@ hint.arge.1.mdio=mdioproxy1 # .. off of hint.ukswitch.0.at=mdio0 hint.ukswitch.0.phymask=0x3 +# Don't flip on anything that isn't already enabled. +# This includes leaving the SPI CS1/CS2 pins as GPIO pins as they're +# not used here. +hint.gpio.0.function_set=0x +hint.gpio.0.function_clear=0x + +# These are the GPIO LEDs and buttons which can be software controlled. +hint.gpio.0.pinmask=0x00ff + +# GPIO 0: Pin 1 +# GPIO 1: Pin 2 +# GPIO 2: RF LED +# GPIO 3: Pin 3 +# GPIO 4: Pin 4 +# GPIO 5: Pin 5 +# GPIO 6: Pin 6 +# GPIO 7: Pin 7 + # RF led hint.gpioled.0.at=gpiobus0 hint.gpioled.0.name=rf ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254990 - head/sys/mips/atheros
Author: loos Date: Wed Aug 28 14:46:15 2013 New Revision: 254990 URL: http://svnweb.freebsd.org/changeset/base/254990 Log: Make ar71xx_spi attach the next free unit of spibus and not only spibus0. Approved by: adrian (mentor) Modified: head/sys/mips/atheros/ar71xx_spi.c Modified: head/sys/mips/atheros/ar71xx_spi.c == --- head/sys/mips/atheros/ar71xx_spi.c Wed Aug 28 14:43:04 2013 (r254989) +++ head/sys/mips/atheros/ar71xx_spi.c Wed Aug 28 14:46:15 2013 (r254990) @@ -108,7 +108,7 @@ ar71xx_spi_attach(device_t dev) SPI_WRITE(sc, AR71XX_SPI_CTRL, 0x43); SPI_WRITE(sc, AR71XX_SPI_IO_CTRL, SPI_IO_CTRL_CSMASK); - device_add_child(dev, spibus, 0); + device_add_child(dev, spibus, -1); return (bus_generic_attach(dev)); } ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254991 - head/sys/dev/flash
Author: loos
Date: Wed Aug 28 14:49:36 2013
New Revision: 254991
URL: http://svnweb.freebsd.org/changeset/base/254991
Log:
Fix a few typos for s25fl types.
Approved by: adrian (mentor)
Modified:
head/sys/dev/flash/mx25l.c
Modified: head/sys/dev/flash/mx25l.c
==
--- head/sys/dev/flash/mx25l.c Wed Aug 28 14:46:15 2013(r254990)
+++ head/sys/dev/flash/mx25l.c Wed Aug 28 14:49:36 2013(r254991)
@@ -104,9 +104,9 @@ struct mx25l_flash_ident flash_devices[]
{ mx25ll32, 0xc2, 0x2016, 64 * 1024, 64, FL_NONE },
{ mx25ll64, 0xc2, 0x2017, 64 * 1024, 128, FL_NONE },
{ mx25ll128, 0xc2, 0x2018, 64 * 1024, 256, FL_ERASE_4K |
FL_ERASE_32K },
+ { s25fl032, 0x01, 0x0215, 64 * 1024, 64, FL_NONE },
+ { s25fl064, 0x01, 0x0216, 64 * 1024, 128, FL_NONE },
{ s25fl128, 0x01, 0x2018, 64 * 1024, 256, FL_NONE },
- { s25s1032, 0x01, 0x0215, 64 * 1024, 64, FL_NONE },
- { s25sl064a, 0x01, 0x0216, 64 * 1024, 128, FL_NONE },
{ SST25VF032B, 0xbf, 0x254a, 64 * 1024, 64, FL_ERASE_4K |
FL_ERASE_32K },
/* Winbond -- w25x blocks are 64K, sectors are 4KiB */
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254992 - head/etc/rc.d
Author: gavin
Date: Wed Aug 28 15:12:15 2013
New Revision: 254992
URL: http://svnweb.freebsd.org/changeset/base/254992
Log:
After writing a kernel core dump into /var/crash, call sync(8).
If we panic again shortly after boot (say, within 30 seconds), any core
dump we wrote out may be lost on reboot. In this situation, we really
want to keep that core file, as it may be the only way to have the issue
resolved. Call sync(8) after writing out the core file and running
crashinfo(8), in the hope that these will not be lost if we panic
again. sync(8) is only called in the case where there is a core dump
to be written out, so won't be called during normal boots.
Discovered by:Trying to debug an IPSEC panic
MFC after:1 week
Modified:
head/etc/rc.d/savecore
Modified: head/etc/rc.d/savecore
==
--- head/etc/rc.d/savecore Wed Aug 28 14:49:36 2013(r254991)
+++ head/etc/rc.d/savecore Wed Aug 28 15:12:15 2013(r254992)
@@ -70,6 +70,7 @@ savecore_start()
if checkyesno crashinfo_enable; then
${crashinfo_program} -d ${dumpdir}
fi
+ sync
else
check_startmsgs echo 'No core dumps found.'
fi
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r254992 - head/etc/rc.d
On Wed, Aug 28, 2013 at 03:12:16PM +, Gavin Atkinson wrote:
Author: gavin
Date: Wed Aug 28 15:12:15 2013
New Revision: 254992
URL: http://svnweb.freebsd.org/changeset/base/254992
Log:
After writing a kernel core dump into /var/crash, call sync(8).
If we panic again shortly after boot (say, within 30 seconds), any core
dump we wrote out may be lost on reboot. In this situation, we really
want to keep that core file, as it may be the only way to have the issue
resolved. Call sync(8) after writing out the core file and running
crashinfo(8), in the hope that these will not be lost if we panic
again. sync(8) is only called in the case where there is a core dump
to be written out, so won't be called during normal boots.
Discovered by: Trying to debug an IPSEC panic
MFC after: 1 week
Modified:
head/etc/rc.d/savecore
Modified: head/etc/rc.d/savecore
==
--- head/etc/rc.d/savecoreWed Aug 28 14:49:36 2013(r254991)
+++ head/etc/rc.d/savecoreWed Aug 28 15:12:15 2013(r254992)
@@ -70,6 +70,7 @@ savecore_start()
if checkyesno crashinfo_enable; then
${crashinfo_program} -d ${dumpdir}
fi
+ sync
else
check_startmsgs echo 'No core dumps found.'
fi
The crashinfo(8) runs are usually quite long, and the program does not
add a new information comparing to the data already present in the
crash dump itself. For your goal, it is more useful to sync before
crashinfo(8) invocation.
pgp5mJFDqG8t5.pgp
Description: PGP signature
Re: svn commit: r254992 - head/etc/rc.d
Gavin Atkinson wrote this message on Wed, Aug 28, 2013 at 15:12 +: Author: gavin Date: Wed Aug 28 15:12:15 2013 New Revision: 254992 URL: http://svnweb.freebsd.org/changeset/base/254992 Log: After writing a kernel core dump into /var/crash, call sync(8). If we panic again shortly after boot (say, within 30 seconds), any core dump we wrote out may be lost on reboot. In this situation, we really want to keep that core file, as it may be the only way to have the issue resolved. Call sync(8) after writing out the core file and running crashinfo(8), in the hope that these will not be lost if we panic again. sync(8) is only called in the case where there is a core dump to be written out, so won't be called during normal boots. Why don't you call fsync on the relevant file(s) (or all the files in the crash dir)? sync just starts the process while fsync will block till all the blocks are on disk... -- John-Mark Gurney Voice: +1 415 225 5579 All that I will do, has been done, All that I have, has not. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254994 - head/lib/msun/src
Author: kargl
Date: Wed Aug 28 16:59:55 2013
New Revision: 254994
URL: http://svnweb.freebsd.org/changeset/base/254994
Log:
* Whitespace.
Modified:
head/lib/msun/src/s_erf.c
Modified: head/lib/msun/src/s_erf.c
==
--- head/lib/msun/src/s_erf.c Wed Aug 28 15:12:51 2013(r254993)
+++ head/lib/msun/src/s_erf.c Wed Aug 28 16:59:55 2013(r254994)
@@ -201,7 +201,7 @@ erf(double x)
if(ix 0x3feb) { /* |x|0.84375 */
if(ix 0x3e30) { /* |x|2**-28 */
if (ix 0x0080)
- return (8*x+efx8*x)/8; /* avoid spurious underflow */
+ return (8*x+efx8*x)/8; /* avoid spurious underflow */
return x + efx*x;
}
z = x*x;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254995 - head/sys/ufs/ffs
Author: mckusick Date: Wed Aug 28 17:38:05 2013 New Revision: 254995 URL: http://svnweb.freebsd.org/changeset/base/254995 Log: A performance problem was reported in PR kern/181226: I have 25TB Dell PERC 6 RAID5 array. When it becomes almost full (10-20GB free), processes which write data to it start eating 100% CPU and write speed drops below 1MB/sec (normally to gives 400MB/sec). The revision at which it first became apparent was http://svnweb.freebsd.org/changeset/base/249782. The offending change reserved an area in each cylinder group to store metadata. The new algorithm attempts to save this area for metadata and allows its use for non-metadata only after all the data areas have been exhausted. The size of the reserved area defaults to half of minfree, so the filesystem reports full before the data area can completely fill. However, in this report, the filesystem has had minfree reduced to 1% thus forcing the metadata area to be used for data. As the filesystem approached full, it had only metadata areas left to allocate. The result was that every block allocation had to scan summary data for 30,000 cylinder groups before falling back to searching up to 30,000 metadata areas. The fix is to give up on saving the metadata areas once the free space reserve drops below 2%. The effect of this change is to use the old algorithm of just accepting the first available block that we find. Since most filesystems use the default 5% minfree, this will have no effect on their operation. For those that want to push to the limit, they will get their crappy block placements quickly. Submitted by: Dmitry Sivachenko Fix Tested by: Dmitry Sivachenko PR:kern/181226 MFC after: 2 weeks Modified: head/sys/ufs/ffs/ffs_alloc.c Modified: head/sys/ufs/ffs/ffs_alloc.c == --- head/sys/ufs/ffs/ffs_alloc.cWed Aug 28 16:59:55 2013 (r254994) +++ head/sys/ufs/ffs/ffs_alloc.cWed Aug 28 17:38:05 2013 (r254995) @@ -516,7 +516,13 @@ ffs_reallocblks_ufs1(ap) ip = VTOI(vp); fs = ip-i_fs; ump = ip-i_ump; - if (fs-fs_contigsumsize = 0) + /* +* If we are not tracking block clusters or if we have less than 2% +* free blocks left, then do not attempt to cluster. Running with +* less than 5% free block reserve is not recommended and those that +* choose to do so do not expect to have good file layout. +*/ + if (fs-fs_contigsumsize = 0 || freespace(fs, 2) 0) return (ENOSPC); buflist = ap-a_buflist; len = buflist-bs_nchildren; @@ -737,7 +743,13 @@ ffs_reallocblks_ufs2(ap) ip = VTOI(vp); fs = ip-i_fs; ump = ip-i_ump; - if (fs-fs_contigsumsize = 0) + /* +* If we are not tracking block clusters or if we have less than 2% +* free blocks left, then do not attempt to cluster. Running with +* less than 5% free block reserve is not recommended and those that +* choose to do so do not expect to have good file layout. +*/ + if (fs-fs_contigsumsize = 0 || freespace(fs, 2) 0) return (ENOSPC); buflist = ap-a_buflist; len = buflist-bs_nchildren; ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254996 - head/sys/ufs/ffs
Author: mckusick
Date: Wed Aug 28 17:46:32 2013
New Revision: 254996
URL: http://svnweb.freebsd.org/changeset/base/254996
Log:
In looking at block layouts as part of fixing filesystem block
allocations under low free-space conditions (-r254995), determine
that old block-preference search order used before -r249782 worked
a bit better. This change reverts to that block-preference search order.
MFC after:2 weeks
Modified:
head/sys/ufs/ffs/ffs_alloc.c
Modified: head/sys/ufs/ffs/ffs_alloc.c
==
--- head/sys/ufs/ffs/ffs_alloc.cWed Aug 28 17:38:05 2013
(r254995)
+++ head/sys/ufs/ffs/ffs_alloc.cWed Aug 28 17:46:32 2013
(r254996)
@@ -1186,7 +1186,7 @@ ffs_dirpref(pip)
if (fs-fs_contigdirs[cg] maxcontigdirs)
return ((ino_t)(fs-fs_ipg * cg));
}
- for (cg = prefcg - 1; cg = 0; cg--)
+ for (cg = 0; cg prefcg; cg++)
if (fs-fs_cs(fs, cg).cs_ndir maxndir
fs-fs_cs(fs, cg).cs_nifree = minifree
fs-fs_cs(fs, cg).cs_nbfree = minbfree) {
@@ -1199,7 +1199,7 @@ ffs_dirpref(pip)
for (cg = prefcg; cg fs-fs_ncg; cg++)
if (fs-fs_cs(fs, cg).cs_nifree = avgifree)
return ((ino_t)(fs-fs_ipg * cg));
- for (cg = prefcg - 1; cg = 0; cg--)
+ for (cg = 0; cg prefcg; cg++)
if (fs-fs_cs(fs, cg).cs_nifree = avgifree)
break;
return ((ino_t)(fs-fs_ipg * cg));
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254997 - head/sys/dev/fb
Author: jkim
Date: Wed Aug 28 17:58:30 2013
New Revision: 254997
URL: http://svnweb.freebsd.org/changeset/base/254997
Log:
Avoid unnecessary signedness conversion.
Modified:
head/sys/dev/fb/vesa.c
Modified: head/sys/dev/fb/vesa.c
==
--- head/sys/dev/fb/vesa.c Wed Aug 28 17:46:32 2013(r254996)
+++ head/sys/dev/fb/vesa.c Wed Aug 28 17:58:30 2013(r254997)
@@ -83,7 +83,7 @@ static struct mtx vesa_lock;
static int vesa_state;
static void *vesa_state_buf;
static uint32_t vesa_state_buf_offs;
-static ssize_t vesa_state_buf_size;
+static size_t vesa_state_buf_size;
static u_char *vesa_palette;
static uint32_t vesa_palette_offs;
@@ -207,7 +207,7 @@ static int vesa_bios_load_palette2(int s
#define STATE_SIZE 0
#define STATE_SAVE 1
#define STATE_LOAD 2
-static ssize_t vesa_bios_state_buf_size(int);
+static size_t vesa_bios_state_buf_size(int);
static int vesa_bios_save_restore(int code, void *p);
#ifdef MODE_TABLE_BROKEN
static int vesa_bios_get_line_length(void);
@@ -505,7 +505,7 @@ vesa_bios_load_palette2(int start, int c
return (regs.R_AX != 0x004f);
}
-static ssize_t
+static size_t
vesa_bios_state_buf_size(int state)
{
x86regs_t regs;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254998 - head/sys/dev/fb
Author: jkim
Date: Wed Aug 28 18:13:37 2013
New Revision: 254998
URL: http://svnweb.freebsd.org/changeset/base/254998
Log:
Make sure to free stale buffer before allocating new one for safety.
Modified:
head/sys/dev/fb/vesa.c
Modified: head/sys/dev/fb/vesa.c
==
--- head/sys/dev/fb/vesa.c Wed Aug 28 17:58:30 2013(r254997)
+++ head/sys/dev/fb/vesa.c Wed Aug 28 18:13:37 2013(r254998)
@@ -1475,6 +1475,10 @@ vesa_save_state(video_adapter_t *adp, vo
if (vesa_state_buf_size 0 size bsize)
return (EINVAL);
+ if (vesa_vmem_buf != NULL) {
+ free(vesa_vmem_buf, M_DEVBUF);
+ vesa_vmem_buf = NULL;
+ }
if (VESA_MODE(adp-va_mode) adp-va_buffer != 0) {
buf = adp-va_buffer;
bsize = adp-va_buffer_size;
@@ -1486,8 +1490,7 @@ vesa_save_state(video_adapter_t *adp, vo
vesa_vmem_buf = malloc(bsize, M_DEVBUF, M_NOWAIT);
if (vesa_vmem_buf != NULL)
bcopy((void *)buf, vesa_vmem_buf, bsize);
- } else
- vesa_vmem_buf = NULL;
+ }
if (vesa_state_buf_size == 0)
return ((*prevvidsw-save_state)(adp, p, size));
((adp_state_t *)p)-sig = V_STATE_SIG;
@@ -1524,6 +1527,7 @@ vesa_load_state(video_adapter_t *adp, vo
bcopy(vesa_vmem_buf, (void *)buf, bsize);
}
free(vesa_vmem_buf, M_DEVBUF);
+ vesa_vmem_buf = NULL;
}
if (((adp_state_t *)p)-sig != V_STATE_SIG)
return ((*prevvidsw-load_state)(adp, p));
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r254999 - head/sys/dev/fb
Author: jkim
Date: Wed Aug 28 19:06:22 2013
New Revision: 254999
URL: http://svnweb.freebsd.org/changeset/base/254999
Log:
Do not save/restore video memory if we are not using linear frame buffer.
Note this partially revert r233896.
Modified:
head/sys/dev/fb/vesa.c
Modified: head/sys/dev/fb/vesa.c
==
--- head/sys/dev/fb/vesa.c Wed Aug 28 18:13:37 2013(r254998)
+++ head/sys/dev/fb/vesa.c Wed Aug 28 19:06:22 2013(r254999)
@@ -1,6 +1,6 @@
/*-
* Copyright (c) 1998 Kazutaka YOKOTA and Michael Smith
- * Copyright (c) 2009-2012 Jung-uk Kim j...@freebsd.org
+ * Copyright (c) 2009-2013 Jung-uk Kim j...@freebsd.org
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -1463,7 +1463,7 @@ vesa_set_border(video_adapter_t *adp, in
static int
vesa_save_state(video_adapter_t *adp, void *p, size_t size)
{
- vm_offset_t buf;
+ void *buf;
size_t bsize;
if (adp != vesa_adp || (size == 0 vesa_state_buf_size == 0))
@@ -1479,17 +1479,14 @@ vesa_save_state(video_adapter_t *adp, vo
free(vesa_vmem_buf, M_DEVBUF);
vesa_vmem_buf = NULL;
}
- if (VESA_MODE(adp-va_mode) adp-va_buffer != 0) {
- buf = adp-va_buffer;
- bsize = adp-va_buffer_size;
- } else {
- buf = adp-va_window;
- bsize = adp-va_window_size;
- }
- if (buf != 0) {
- vesa_vmem_buf = malloc(bsize, M_DEVBUF, M_NOWAIT);
- if (vesa_vmem_buf != NULL)
- bcopy((void *)buf, vesa_vmem_buf, bsize);
+ if (VESA_MODE(adp-va_mode)) {
+ buf = (void *)adp-va_buffer;
+ if (buf != NULL) {
+ bsize = adp-va_buffer_size;
+ vesa_vmem_buf = malloc(bsize, M_DEVBUF, M_NOWAIT);
+ if (vesa_vmem_buf != NULL)
+ bcopy(buf, vesa_vmem_buf, bsize);
+ }
}
if (vesa_state_buf_size == 0)
return ((*prevvidsw-save_state)(adp, p, size));
@@ -1500,7 +1497,7 @@ vesa_save_state(video_adapter_t *adp, vo
static int
vesa_load_state(video_adapter_t *adp, void *p)
{
- vm_offset_t buf;
+ void *buf;
size_t bsize;
int error, mode;
@@ -1515,16 +1512,12 @@ vesa_load_state(video_adapter_t *adp, vo
error = vesa_set_mode(adp, mode);
if (vesa_vmem_buf != NULL) {
- if (error == 0) {
- if (VESA_MODE(mode) adp-va_buffer != 0) {
- buf = adp-va_buffer;
+ if (error == 0 VESA_MODE(mode)) {
+ buf = (void *)adp-va_buffer;
+ if (buf != NULL) {
bsize = adp-va_buffer_size;
- } else {
- buf = adp-va_window;
- bsize = adp-va_window_size;
+ bcopy(vesa_vmem_buf, buf, bsize);
}
- if (buf != 0)
- bcopy(vesa_vmem_buf, (void *)buf, bsize);
}
free(vesa_vmem_buf, M_DEVBUF);
vesa_vmem_buf = NULL;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255000 - stable/9/sys/cam/scsi
Author: ken
Date: Wed Aug 28 19:22:09 2013
New Revision: 255000
URL: http://svnweb.freebsd.org/changeset/base/255000
Log:
MFC 254416:
r254416 | ken | 2013-08-16 10:14:32 -0600 (Fri, 16 Aug 2013) | 24 lines
Add unmapped I/O and larger I/O support to the sa(4) driver.
We now pay attention to the maxio field in the XPT_PATH_INQ CCB,
and if it is set, propagate it up to physio via the si_iosize_max
field in the cdev structure.
We also now pay attention to the PIM_UNMAPPED capability bit in the
XPT_PATH_INQ CCB, and set the new SI_UNMAPPED cdev flag when the
underlying SIM supports unmapped I/O.
scsi_sa.c: Add unmapped I/O support and propagate the SIM's
maximum I/O size up.
Adjust scsi_tape_read_write() in the same way that
scsi_read_write() was changed to support unmapped
I/O. We overload the readop parameter with bits
that tell us whether it's an unmapped I/O, and we
need to set the CAM_DATA_BIO CCB flag. This change
should be backwards compatible in source and
binary forms.
Sponsored by: Spectra Logic
Modified:
stable/9/sys/cam/scsi/scsi_sa.c
Directory Properties:
stable/9/sys/ (props changed)
Modified: stable/9/sys/cam/scsi/scsi_sa.c
==
--- stable/9/sys/cam/scsi/scsi_sa.c Wed Aug 28 19:06:22 2013
(r254999)
+++ stable/9/sys/cam/scsi/scsi_sa.c Wed Aug 28 19:22:09 2013
(r255000)
@@ -212,6 +212,7 @@ struct sa_softc {
sa_statestate;
sa_flagsflags;
sa_quirks quirks;
+ u_int si_flags;
struct bio_queue_head bio_queue;
int queue_count;
struct devstat *device_stats;
@@ -221,6 +222,7 @@ struct sa_softc {
int blk_shift;
u_int32_t max_blk;
u_int32_t min_blk;
+ u_int32_t maxio;
u_int32_t comp_algorithm;
u_int32_t saved_comp_algorithm;
u_int32_t media_blksize;
@@ -447,7 +449,7 @@ static struct cdevsw sa_cdevsw = {
.d_ioctl = saioctl,
.d_strategy = sastrategy,
.d_name = sa,
- .d_flags = D_TAPE | D_NEEDGIANT,
+ .d_flags = D_TAPE,
};
static int
@@ -1506,10 +1508,29 @@ saregister(struct cam_periph *periph, vo
DEVSTAT_BS_UNAVAILABLE, SID_TYPE(cgd-inq_data) |
XPORT_DEVSTAT_TYPE(cpi.transport), DEVSTAT_PRIORITY_TAPE);
+ /*
+* If maxio isn't set, we fall back to DFLTPHYS. If it is set, we
+* take it whether or not it's larger than MAXPHYS. physio will
+* break it down into pieces small enough to fit in a buffer.
+*/
+ if (cpi.maxio == 0)
+ softc-maxio = DFLTPHYS;
+ else
+ softc-maxio = cpi.maxio;
+
+ /*
+* If the SIM supports unmapped I/O, let physio know that we can
+* handle unmapped buffers.
+*/
+ if (cpi.hba_misc PIM_UNMAPPED)
+ softc-si_flags = SI_UNMAPPED;
+
softc-devs.ctl_dev = make_dev(sa_cdevsw, SAMINOR(SA_CTLDEV,
0, SA_ATYPE_R), UID_ROOT, GID_OPERATOR,
0660, %s%d.ctl, periph-periph_name, periph-unit_number);
softc-devs.ctl_dev-si_drv1 = periph;
+ softc-devs.ctl_dev-si_iosize_max = softc-maxio;
+ softc-devs.ctl_dev-si_flags |= softc-si_flags;
for (i = 0; i SA_NUM_MODES; i++) {
@@ -1518,18 +1539,24 @@ saregister(struct cam_periph *periph, vo
UID_ROOT, GID_OPERATOR, 0660, %s%d.%d,
periph-periph_name, periph-unit_number, i);
softc-devs.mode_devs[i].r_dev-si_drv1 = periph;
+ softc-devs.mode_devs[i].r_dev-si_iosize_max = softc-maxio;
+ softc-devs.mode_devs[i].r_dev-si_flags |= softc-si_flags;
softc-devs.mode_devs[i].nr_dev = make_dev(sa_cdevsw,
SAMINOR(SA_NOT_CTLDEV, i, SA_ATYPE_NR),
UID_ROOT, GID_OPERATOR, 0660, n%s%d.%d,
periph-periph_name, periph-unit_number, i);
softc-devs.mode_devs[i].nr_dev-si_drv1 = periph;
+ softc-devs.mode_devs[i].nr_dev-si_iosize_max = softc-maxio;
+ softc-devs.mode_devs[i].nr_dev-si_flags |= softc-si_flags;
softc-devs.mode_devs[i].er_dev = make_dev(sa_cdevsw,
SAMINOR(SA_NOT_CTLDEV, i, SA_ATYPE_ER),
UID_ROOT, GID_OPERATOR, 0660, e%s%d.%d,
periph-periph_name, periph-unit_number, i);
softc-devs.mode_devs[i].er_dev-si_drv1 = periph;
+ softc-devs.mode_devs[i].er_dev-si_iosize_max = softc-maxio;
+
svn commit: r255001 - head/share/man/man4
Author: rwatson Date: Wed Aug 28 19:49:32 2013 New Revision: 255001 URL: http://svnweb.freebsd.org/changeset/base/255001 Log: Add a simple procdesc(4) man page describing options PROCDESC and the high-level facility, supplementing pdfork(2) and friends. Update capsicum.4 to xref. Suggested by: sbruno MFC after:3 days Added: head/share/man/man4/procdesc.4 (contents, props changed) Modified: head/share/man/man4/Makefile head/share/man/man4/capsicum.4 Modified: head/share/man/man4/Makefile == --- head/share/man/man4/MakefileWed Aug 28 19:22:09 2013 (r255000) +++ head/share/man/man4/MakefileWed Aug 28 19:49:32 2013 (r255001) @@ -367,6 +367,7 @@ MAN=aac.4 \ ppbus.4 \ ppc.4 \ ppi.4 \ + procdesc.4 \ psm.4 \ pst.4 \ pt.4 \ Modified: head/share/man/man4/capsicum.4 == --- head/share/man/man4/capsicum.4 Wed Aug 28 19:22:09 2013 (r255000) +++ head/share/man/man4/capsicum.4 Wed Aug 28 19:49:32 2013 (r255001) @@ -1,5 +1,5 @@ .\ -.\ Copyright (c) 2011 Robert N. M. Watson +.\ Copyright (c) 2011, 2013 Robert N. M. Watson .\ Copyright (c) 2011 Jonathan Anderson .\ All rights reserved. .\ @@ -26,7 +26,7 @@ .\ .\ $FreeBSD$ .\ -.Dd September 20, 2011 +.Dd August 21, 2013 .Dt CAPSICUM 4 .Os .Sh NAME @@ -78,10 +78,13 @@ objects using capabilities rather than g .Bl -tag -width indent .It process descriptors File descriptors representing processes, allowing parent processes to manage -child processes without requiring access to the PID namespace. +child processes without requiring access to the PID namespace; described in +greater detail in +.Xr procdesc 4 . .It anonymous shared memory An extension to the POSIX shared memory API to support anonymous swap objects -associated with file descriptors. +associated with file descriptors; described in greater detail in +.Xr shm_open 2 . .El .Sh SEE ALSO .Xr cap_enter 2 , @@ -96,7 +99,8 @@ associated with file descriptors. .Xr pdwait4 2 , .Xr read 2 , .Xr shm_open 2 , -.Xr write 2 +.Xr write 2 , +.Xr procdesc 4 , .Sh HISTORY .Nm first appeared in Added: head/share/man/man4/procdesc.4 == --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/share/man/man4/procdesc.4 Wed Aug 28 19:49:32 2013 (r255001) @@ -0,0 +1,93 @@ +.\ +.\ Copyright (c) 2013 Robert N. M. Watson +.\ All rights reserved. +.\ +.\ This software was developed by SRI International and the University of +.\ Cambridge Computer Laboratory under DARPA/AFRL contract (FA8750-10-C-0237) +.\ (CTSRD), as part of the DARPA CRASH research programme. +.\ +.\ 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. +.\ +.\ $FreeBSD$ +.\ +.Dd August 21, 2013 +.Dt PROCDESC 4 +.Os +.Sh NAME +.Nm procdesc +.Nd process descriptor facility +.Sh SYNOPSIS +.Cd options PROCDESC +.Sh DESCRIPTION +.Nm +is a file-descriptor-oriented interface to process signalling and control, +which supplements historic +.Ux +.Xr fork 2 , +.Xr kill 2 , +and +.Xr wait4 2 +primitives with +new system calls such as +.Xr pdfork 2 , +.Xr pdkill 2 , +and +.Xr pdwait4 2 . +.Nm +is designed for use with +.Xr capsicum 4 , +replacing process identifiers with capability-oriented references. +However, it can also be used independently of +.Xr capsicum 4 , +displacing PIDs, which may otherwise suffer from race conditions. +Given a process descriptor, it is possible to query its conventional PID using +.Xr pdgetpid 2 . +.Sh SEE
svn commit: r255002 - head/lib/libc/sys
Author: rwatson Date: Wed Aug 28 20:00:25 2013 New Revision: 255002 URL: http://svnweb.freebsd.org/changeset/base/255002 Log: Xref capsicum(4) and procdesc(4) from pdfork(2). Suggested by: sbruno MFC after:3 days Modified: head/lib/libc/sys/pdfork.2 Modified: head/lib/libc/sys/pdfork.2 == --- head/lib/libc/sys/pdfork.2 Wed Aug 28 19:49:32 2013(r255001) +++ head/lib/libc/sys/pdfork.2 Wed Aug 28 20:00:25 2013(r255002) @@ -1,10 +1,14 @@ .\ -.\ Copyright (c) 2009-2010, 2012 Robert N. M. Watson +.\ Copyright (c) 2009-2010, 2012-2013 Robert N. M. Watson .\ All rights reserved. .\ .\ This software was developed at the University of Cambridge Computer .\ Laboratory with support from a grant from Google, Inc. .\ +.\ This software was developed by SRI International and the University of +.\ Cambridge Computer Laboratory under DARPA/AFRL contract (FA8750-10-C-0237) +.\ (CTSRD), as part of the DARPA CRASH research programme. +.\ .\ Redistribution and use in source and binary forms, with or without .\ modification, are permitted provided that the following conditions .\ are met: @@ -28,7 +32,7 @@ .\ .\ $FreeBSD$ .\ -.Dd February 25, 2012 +.Dd August 28, 2013 .Dt PDFORK 2 .Os .Sh NAME @@ -71,7 +75,9 @@ Instead of the default terminate-on-clos live until it is explicitly killed with .Xr kill 2 . .Pp -This option is not permitted in Capsicum capability mode (see +This option is not permitted in +.Xr capsicum 4 +capability mode (see .Xr cap_enter 2 ) . .El .Pp @@ -119,6 +125,12 @@ is set; if the process is still alive an the last reference to the process descriptor, the process will be terminated with the signal .Dv SIGKILL . +.Pp +.Nm +and associated functions depend on +.Cd options PROCDESC +described in +.Xr procdesc 4 . .Sh RETURN VALUES .Fn pdfork returns a PID, 0 or -1, as @@ -156,7 +168,9 @@ for .Xr fstat 2 , .Xr kill 2 , .Xr poll 2 , -.Xr wait4 2 +.Xr wait4 2 , +.Xr capsicum 4 , +.Xr procdesc 4 .Sh HISTORY The .Fn pdfork , ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255003 - head/sys/dev/qlxgbe
Author: davidcs
Date: Wed Aug 28 20:07:00 2013
New Revision: 255003
URL: http://svnweb.freebsd.org/changeset/base/255003
Log:
ql_minidump() should be performed only by port 0.
Submitted by: David C Somayajulu
Modified:
head/sys/dev/qlxgbe/ql_os.c
Modified: head/sys/dev/qlxgbe/ql_os.c
==
--- head/sys/dev/qlxgbe/ql_os.c Wed Aug 28 20:00:25 2013(r255002)
+++ head/sys/dev/qlxgbe/ql_os.c Wed Aug 28 20:07:00 2013(r255003)
@@ -1642,8 +1642,6 @@ qla_error_recovery(void *context, int pe
QLA_UNLOCK(ha, __func__);
- ql_minidump(ha);
-
if ((ha-pci_func 0x1) == 0) {
if (!ha-msg_from_peer) {
@@ -1656,6 +1654,8 @@ qla_error_recovery(void *context, int pe
ha-msg_from_peer = 0;
+ ql_minidump(ha);
+
(void) ql_init_hw(ha);
qla_free_xmt_bufs(ha);
qla_free_rcv_bufs(ha);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255004 - head/sys/dev/fb
Author: jkim
Date: Wed Aug 28 20:10:56 2013
New Revision: 255004
URL: http://svnweb.freebsd.org/changeset/base/255004
Log:
Reduce diff against stable/9 slightly.
Modified:
head/sys/dev/fb/vesa.c
Modified: head/sys/dev/fb/vesa.c
==
--- head/sys/dev/fb/vesa.c Wed Aug 28 20:07:00 2013(r255003)
+++ head/sys/dev/fb/vesa.c Wed Aug 28 20:10:56 2013(r255004)
@@ -1506,6 +1506,7 @@ vesa_load_state(video_adapter_t *adp, vo
/* Try BIOS POST to restore a sane state. */
(void)vesa_bios_post();
+ bsize = adp-va_buffer_size;
mode = adp-va_mode;
error = vesa_set_mode(adp, adp-va_initial_mode);
if (mode != adp-va_initial_mode)
@@ -1514,10 +1515,8 @@ vesa_load_state(video_adapter_t *adp, vo
if (vesa_vmem_buf != NULL) {
if (error == 0 VESA_MODE(mode)) {
buf = (void *)adp-va_buffer;
- if (buf != NULL) {
- bsize = adp-va_buffer_size;
+ if (buf != NULL)
bcopy(vesa_vmem_buf, buf, bsize);
- }
}
free(vesa_vmem_buf, M_DEVBUF);
vesa_vmem_buf = NULL;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255005 - in head/sys/dev/cxgbe: . tom
Author: np
Date: Wed Aug 28 20:45:45 2013
New Revision: 255005
URL: http://svnweb.freebsd.org/changeset/base/255005
Log:
Add hooks in base cxgbe(4) for the iWARP upper-layer driver. Update a
couple of assertions in the TOE driver as well.
Modified:
head/sys/dev/cxgbe/adapter.h
head/sys/dev/cxgbe/offload.h
head/sys/dev/cxgbe/osdep.h
head/sys/dev/cxgbe/t4_main.c
head/sys/dev/cxgbe/t4_sge.c
head/sys/dev/cxgbe/tom/t4_cpl_io.c
Modified: head/sys/dev/cxgbe/adapter.h
==
--- head/sys/dev/cxgbe/adapter.hWed Aug 28 20:10:56 2013
(r255004)
+++ head/sys/dev/cxgbe/adapter.hWed Aug 28 20:45:45 2013
(r255005)
@@ -567,6 +567,7 @@ struct adapter {
#ifdef TCP_OFFLOAD
void *tom_softc;/* (struct tom_data *) */
struct tom_tunables tt;
+ void *iwarp_softc; /* (struct c4iw_dev *) */
#endif
struct l2t_data *l2t; /* L2 table */
struct tid_info tids;
Modified: head/sys/dev/cxgbe/offload.h
==
--- head/sys/dev/cxgbe/offload.hWed Aug 28 20:10:56 2013
(r255004)
+++ head/sys/dev/cxgbe/offload.hWed Aug 28 20:45:45 2013
(r255005)
@@ -123,6 +123,7 @@ struct t4_virt_res {
#ifdef TCP_OFFLOAD
enum {
ULD_TOM = 1,
+ ULD_IWARP = 2,
};
struct adapter;
Modified: head/sys/dev/cxgbe/osdep.h
==
--- head/sys/dev/cxgbe/osdep.h Wed Aug 28 20:10:56 2013(r255004)
+++ head/sys/dev/cxgbe/osdep.h Wed Aug 28 20:45:45 2013(r255005)
@@ -45,6 +45,7 @@
#define CH_ALERT(adap, fmt, ...) log(LOG_ALERT, fmt, ##__VA_ARGS__)
#define CH_WARN_RATELIMIT(adap, fmt, ...) log(LOG_WARNING, fmt, ##__VA_ARGS__)
+#ifndef LINUX_TYPES_DEFINED
typedef int8_t s8;
typedef int16_t s16;
typedef int32_t s32;
@@ -156,5 +157,6 @@ strstrip(char *s)
return (r);
}
+#endif /* LINUX_TYPES_DEFINED */
#endif
Modified: head/sys/dev/cxgbe/t4_main.c
==
--- head/sys/dev/cxgbe/t4_main.cWed Aug 28 20:10:56 2013
(r255004)
+++ head/sys/dev/cxgbe/t4_main.cWed Aug 28 20:45:45 2013
(r255005)
@@ -600,7 +600,6 @@ t4_attach(device_t dev)
t4_register_cpl_handler(sc, CPL_TRACE_PKT_T5, t5_trace_pkt);
t4_init_sge_cpl_handlers(sc);
-
/* Prepare the adapter for operation */
rc = -t4_prep_adapter(sc);
if (rc != 0) {
Modified: head/sys/dev/cxgbe/t4_sge.c
==
--- head/sys/dev/cxgbe/t4_sge.c Wed Aug 28 20:10:56 2013(r255004)
+++ head/sys/dev/cxgbe/t4_sge.c Wed Aug 28 20:45:45 2013(r255005)
@@ -1069,6 +1069,17 @@ service_iq(struct sge_iq *iq, int budget
(%s: budget %u, rsp_type %u, __func__,
budget, rsp_type));
+ /*
+* There are 1K interrupt-capable queues (qids 0
+* through 1023). A response type indicating a
+* forwarded interrupt with a qid = 1K is an
+* iWARP async notification.
+*/
+ if (lq = 1024) {
+sc-an_handler(iq, ctrl);
+break;
+}
+
q = sc-sge.iqmap[lq - sc-sge.iq_start];
if (atomic_cmpset_int(q-state, IQS_IDLE,
IQS_BUSY)) {
@@ -1083,7 +1094,12 @@ service_iq(struct sge_iq *iq, int budget
break;
default:
- sc-an_handler(iq, ctrl);
+ KASSERT(0,
+ (%s: illegal response type %d on iq %p,
+ __func__, rsp_type, iq));
+ log(LOG_ERR,
+ %s: illegal response type %d on iq %p,
+ device_get_nameunit(sc-dev), rsp_type, iq);
break;
}
Modified: head/sys/dev/cxgbe/tom/t4_cpl_io.c
==
--- head/sys/dev/cxgbe/tom/t4_cpl_io.c Wed Aug 28 20:10:56 2013
(r255004)
+++ head/sys/dev/cxgbe/tom/t4_cpl_io.c Wed Aug 28 20:45:45 2013
(r255005)
@@ -548,9 +548,10 @@ t4_push_frames(struct adapter *sc, struc
KASSERT(toep-flags TPF_FLOWC_WR_SENT,
(%s: flowc_wr not sent for tid %u., __func__, toep-tid));
- if
Re: svn commit: r254627 - in head: bin/chflags bin/ls lib/libc/gen lib/libc/sys sys/cddl/contrib/opensolaris/uts/common/fs/zfs sys/fs/msdosfs sys/fs/smbfs sys/sys sys/ufs/ufs
On 2013-08-21 18:04, Kenneth D. Merry wrote:
Author: ken
Date: Wed Aug 21 23:04:48 2013
New Revision: 254627
URL: http://svnweb.freebsd.org/changeset/base/254627
Log:
Expand the use of stat(2) flags to allow storing some Windows/DOS
and CIFS file attributes as BSD stat(2) flags.
This work is intended to be compatible with ZFS, the Solaris CIFS
server's interaction with ZFS, somewhat compatible with MacOS X,
and of course compatible with Windows.
The Windows attributes that are implemented were chosen based on
the attributes that ZFS already supports.
The summary of the flags is as follows:
[...]
UF_ARCHIVE: Command line name: uarch, uarchive
ZFS_NAME: XAT_ARCHIVE, ZFS_ARCHIVE
Windows name: FILE_ATTRIBUTE_ARCHIVE
The UF_ARCHIVED flag means that the file has changed and
needs to be archived. The meaning is same as
the Windows FILE_ATTRIBUTE_ARCHIVE attribute, and
the ZFS XAT_ARCHIVE and ZFS_ARCHIVE attribute.
msdosfs and ZFS have special handling for this flag.
i.e. they will set it when the file changes.
Is it intended that this flag is automatically added to all new and
existing ZFS files?
# touch test
# ls -alo test
-rw-r--r--- 1 root wheel uarch 0 Aug 28 15:46 test
This breaks 'cp -p' to tmpfs as tmpfs does not allow this flag.
# mkdir /tmp/tmpfs
# mount -t tmpfs tmpfs /tmp/tmpfs
# cp -f test /tmp/tmpfs
cp: test: Operation not supported
[...]
Modified: head/sys/ufs/ufs/ufs_vnops.c
==
--- head/sys/ufs/ufs/ufs_vnops.cWed Aug 21 22:57:29 2013
(r254626)
+++ head/sys/ufs/ufs/ufs_vnops.cWed Aug 21 23:04:48 2013
(r254627)
@@ -528,9 +528,11 @@ ufs_setattr(ap)
return (EINVAL);
}
if (vap-va_flags != VNOVAL) {
- if ((vap-va_flags ~(UF_NODUMP | UF_IMMUTABLE | UF_APPEND |
- UF_OPAQUE | UF_NOUNLINK | SF_ARCHIVED | SF_IMMUTABLE |
- SF_APPEND | SF_NOUNLINK | SF_SNAPSHOT)) != 0)
+ if ((vap-va_flags ~(SF_APPEND | SF_ARCHIVED | SF_IMMUTABLE |
+ SF_NOUNLINK | SF_SNAPSHOT | UF_APPEND | UF_ARCHIVE |
+ UF_HIDDEN | UF_IMMUTABLE | UF_NODUMP | UF_NOUNLINK |
+ UF_OFFLINE | UF_OPAQUE | UF_READONLY | UF_REPARSE |
+ UF_SPARSE | UF_SYSTEM)) != 0)
return (EOPNOTSUPP);
if (vp-v_mount-mnt_flag MNT_RDONLY)
return (EROFS);
Seems a similar change is needed in tmpfs_subr.c:tmpfs_chflags()
(antoine pointed this out)
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255006 - in head/sys/dev/cxgbe: . tom
Author: np
Date: Wed Aug 28 20:59:22 2013
New Revision: 255006
URL: http://svnweb.freebsd.org/changeset/base/255006
Log:
Change t4_list_lock and t4_uld_list_lock from mutexes to sx'es.
- tom_uninit had to be reworked not to hold the adapter lock (a mutex)
around t4_deactivate_uld, which acquires the uld_list_lock.
- the ifc_match for the interface cloner that creates the tracer ifnet
had to be reworked as the kernel calls ifc_match with the global
if_cloners_mtx held.
Modified:
head/sys/dev/cxgbe/t4_main.c
head/sys/dev/cxgbe/t4_tracer.c
head/sys/dev/cxgbe/tom/t4_tom.c
Modified: head/sys/dev/cxgbe/t4_main.c
==
--- head/sys/dev/cxgbe/t4_main.cWed Aug 28 20:45:45 2013
(r255005)
+++ head/sys/dev/cxgbe/t4_main.cWed Aug 28 20:59:22 2013
(r255006)
@@ -160,10 +160,10 @@ MALLOC_DEFINE(M_CXGBE, cxgbe, Chelsio
* Correct lock order when you need to acquire multiple locks is t4_list_lock,
* then ADAPTER_LOCK, then t4_uld_list_lock.
*/
-static struct mtx t4_list_lock;
+static struct sx t4_list_lock;
static SLIST_HEAD(, adapter) t4_list;
#ifdef TCP_OFFLOAD
-static struct mtx t4_uld_list_lock;
+static struct sx t4_uld_list_lock;
static SLIST_HEAD(, uld_info) t4_uld_list;
#endif
@@ -568,9 +568,9 @@ t4_attach(device_t dev)
snprintf(sc-lockname, sizeof(sc-lockname), %s,
device_get_nameunit(dev));
mtx_init(sc-sc_lock, sc-lockname, 0, MTX_DEF);
- mtx_lock(t4_list_lock);
+ sx_xlock(t4_list_lock);
SLIST_INSERT_HEAD(t4_list, sc, link);
- mtx_unlock(t4_list_lock);
+ sx_xunlock(t4_list_lock);
mtx_init(sc-sfl_lock, starving freelists, 0, MTX_DEF);
TAILQ_INIT(sc-sfl);
@@ -917,9 +917,9 @@ t4_detach(device_t dev)
free(sc-tids.ftid_tab, M_CXGBE);
t4_destroy_dma_tag(sc);
if (mtx_initialized(sc-sc_lock)) {
- mtx_lock(t4_list_lock);
+ sx_xlock(t4_list_lock);
SLIST_REMOVE(t4_list, sc, adapter, link);
- mtx_unlock(t4_list_lock);
+ sx_xunlock(t4_list_lock);
mtx_destroy(sc-sc_lock);
}
@@ -7341,7 +7341,7 @@ t4_iterate(void (*func)(struct adapter *
{
struct adapter *sc;
- mtx_lock(t4_list_lock);
+ sx_slock(t4_list_lock);
SLIST_FOREACH(sc, t4_list, link) {
/*
* func should not make any assumptions about what state sc is
@@ -7349,7 +7349,7 @@ t4_iterate(void (*func)(struct adapter *
*/
func(sc, arg);
}
- mtx_unlock(t4_list_lock);
+ sx_sunlock(t4_list_lock);
}
static int
@@ -7577,7 +7577,7 @@ t4_register_uld(struct uld_info *ui)
int rc = 0;
struct uld_info *u;
- mtx_lock(t4_uld_list_lock);
+ sx_xlock(t4_uld_list_lock);
SLIST_FOREACH(u, t4_uld_list, link) {
if (u-uld_id == ui-uld_id) {
rc = EEXIST;
@@ -7588,7 +7588,7 @@ t4_register_uld(struct uld_info *ui)
SLIST_INSERT_HEAD(t4_uld_list, ui, link);
ui-refcount = 0;
done:
- mtx_unlock(t4_uld_list_lock);
+ sx_xunlock(t4_uld_list_lock);
return (rc);
}
@@ -7598,7 +7598,7 @@ t4_unregister_uld(struct uld_info *ui)
int rc = EINVAL;
struct uld_info *u;
- mtx_lock(t4_uld_list_lock);
+ sx_xlock(t4_uld_list_lock);
SLIST_FOREACH(u, t4_uld_list, link) {
if (u == ui) {
@@ -7613,7 +7613,7 @@ t4_unregister_uld(struct uld_info *ui)
}
}
done:
- mtx_unlock(t4_uld_list_lock);
+ sx_xunlock(t4_uld_list_lock);
return (rc);
}
@@ -7625,7 +7625,7 @@ t4_activate_uld(struct adapter *sc, int
ASSERT_SYNCHRONIZED_OP(sc);
- mtx_lock(t4_uld_list_lock);
+ sx_slock(t4_uld_list_lock);
SLIST_FOREACH(ui, t4_uld_list, link) {
if (ui-uld_id == id) {
@@ -7636,7 +7636,7 @@ t4_activate_uld(struct adapter *sc, int
}
}
done:
- mtx_unlock(t4_uld_list_lock);
+ sx_sunlock(t4_uld_list_lock);
return (rc);
}
@@ -7649,7 +7649,7 @@ t4_deactivate_uld(struct adapter *sc, in
ASSERT_SYNCHRONIZED_OP(sc);
- mtx_lock(t4_uld_list_lock);
+ sx_slock(t4_uld_list_lock);
SLIST_FOREACH(ui, t4_uld_list, link) {
if (ui-uld_id == id) {
@@ -7660,7 +7660,7 @@ t4_deactivate_uld(struct adapter *sc, in
}
}
done:
- mtx_unlock(t4_uld_list_lock);
+ sx_sunlock(t4_uld_list_lock);
return (rc);
}
@@ -7741,10 +7741,10 @@ mod_event(module_t mod, int cmd, void *a
if (atomic_fetchadd_int(loaded, 1))
break;
t4_sge_modload();
- mtx_init(t4_list_lock, T4 adapters, 0, MTX_DEF);
+ sx_init(t4_list_lock, T4/T5 adapters);
svn commit: r255007 - head/lib/libutil
Author: jilles
Date: Wed Aug 28 21:10:37 2013
New Revision: 255007
URL: http://svnweb.freebsd.org/changeset/base/255007
Log:
libutil: Use O_CLOEXEC for internal file descriptors from open().
Modified:
head/lib/libutil/login_auth.c
head/lib/libutil/login_cap.c
head/lib/libutil/pidfile.c
head/lib/libutil/quotafile.c
head/lib/libutil/uucplock.c
Modified: head/lib/libutil/login_auth.c
==
--- head/lib/libutil/login_auth.c Wed Aug 28 20:59:22 2013
(r255006)
+++ head/lib/libutil/login_auth.c Wed Aug 28 21:10:37 2013
(r255007)
@@ -98,7 +98,7 @@ auth_cat(const char *file)
int fd, count;
char buf[BUFSIZ];
- if ((fd = open(file, O_RDONLY)) 0)
+ if ((fd = open(file, O_RDONLY | O_CLOEXEC)) 0)
return 0;
while ((count = read(fd, buf, sizeof(buf))) 0)
(void)write(fileno(stdout), buf, count);
Modified: head/lib/libutil/login_cap.c
==
--- head/lib/libutil/login_cap.cWed Aug 28 20:59:22 2013
(r255006)
+++ head/lib/libutil/login_cap.cWed Aug 28 21:10:37 2013
(r255007)
@@ -239,7 +239,7 @@ login_getclassbyname(char const *name, c
break; /* Don't retry default on 'me' */
if (i == 0)
r = -1;
- else if ((r = open(login_dbarray[0], O_RDONLY)) = 0)
+ else if ((r = open(login_dbarray[0], O_RDONLY | O_CLOEXEC)) = 0)
close(r);
/*
* If there's at least one login class database,
Modified: head/lib/libutil/pidfile.c
==
--- head/lib/libutil/pidfile.c Wed Aug 28 20:59:22 2013(r255006)
+++ head/lib/libutil/pidfile.c Wed Aug 28 21:10:37 2013(r255007)
@@ -73,7 +73,7 @@ pidfile_read(const char *path, pid_t *pi
char buf[16], *endptr;
int error, fd, i;
- fd = open(path, O_RDONLY);
+ fd = open(path, O_RDONLY | O_CLOEXEC);
if (fd == -1)
return (errno);
Modified: head/lib/libutil/quotafile.c
==
--- head/lib/libutil/quotafile.cWed Aug 28 20:59:22 2013
(r255006)
+++ head/lib/libutil/quotafile.cWed Aug 28 21:10:37 2013
(r255007)
@@ -137,7 +137,7 @@ quota_open(struct fstab *fs, int quotaty
goto error;
}
qf-accmode = openflags O_ACCMODE;
- if ((qf-fd = open(qf-qfname, qf-accmode)) 0
+ if ((qf-fd = open(qf-qfname, qf-accmode|O_CLOEXEC)) 0
(openflags O_CREAT) != O_CREAT)
goto error;
/* File open worked, so process it */
@@ -168,7 +168,8 @@ quota_open(struct fstab *fs, int quotaty
/* not reached */
}
/* open failed, but O_CREAT was specified, so create a new file */
- if ((qf-fd = open(qf-qfname, O_RDWR|O_CREAT|O_TRUNC, 0)) 0)
+ if ((qf-fd = open(qf-qfname, O_RDWR|O_CREAT|O_TRUNC|O_CLOEXEC, 0))
+ 0)
goto error;
qf-wordsize = 64;
memset(dqh, 0, sizeof(dqh));
@@ -534,7 +535,8 @@ quota_convert(struct quotafile *qf, int
free(newqf);
return (-1);
}
- if ((newqf-fd = open(qf-qfname, O_RDWR|O_CREAT|O_TRUNC, 0)) 0) {
+ if ((newqf-fd = open(qf-qfname, O_RDWR|O_CREAT|O_TRUNC|O_CLOEXEC,
+ 0)) 0) {
serrno = errno;
goto error;
}
Modified: head/lib/libutil/uucplock.c
==
--- head/lib/libutil/uucplock.c Wed Aug 28 20:59:22 2013(r255006)
+++ head/lib/libutil/uucplock.c Wed Aug 28 21:10:37 2013(r255007)
@@ -76,7 +76,8 @@ uu_lock(const char *tty_name)
pid);
(void)snprintf(lckname, sizeof(lckname), _PATH_UUCPLOCK LOCKFMT,
tty_name);
- if ((tmpfd = creat(lcktmpname, 0664)) 0)
+ if ((tmpfd = open(lcktmpname, O_CREAT | O_TRUNC | O_WRONLY | O_CLOEXEC,
+ 0664)) 0)
GORET(0, UU_LOCK_CREAT_ERR);
for (i = 0; i MAXTRIES; i++) {
@@ -88,7 +89,7 @@ uu_lock(const char *tty_name)
* check to see if the process holding the lock
* still exists
*/
- if ((fd = open(lckname, O_RDONLY)) 0)
+ if ((fd = open(lckname, O_RDONLY | O_CLOEXEC)) 0)
GORET(1, UU_LOCK_OPEN_ERR);
if ((pid_old = get_pid (fd, err)) == -1)
@@ -132,7 +133,7 @@ uu_lock_txfr(const char *tty_name, pid_t
snprintf(lckname, sizeof(lckname), _PATH_UUCPLOCK LOCKFMT, tty_name);
- if ((fd = open(lckname, O_RDWR)) 0)
+ if ((fd = open(lckname, O_RDWR | O_CLOEXEC))
Re: svn commit: r254627 - in head: bin/chflags bin/ls lib/libc/gen lib/libc/sys sys/cddl/contrib/opensolaris/uts/common/fs/zfs sys/fs/msdosfs sys/fs/smbfs sys/sys sys/ufs/ufs
On Wed, Aug 28, 2013 at 15:54:18 -0500, Bryan Drewery wrote:
On 2013-08-21 18:04, Kenneth D. Merry wrote:
Author: ken
Date: Wed Aug 21 23:04:48 2013
New Revision: 254627
URL: http://svnweb.freebsd.org/changeset/base/254627
Log:
Expand the use of stat(2) flags to allow storing some Windows/DOS
and CIFS file attributes as BSD stat(2) flags.
This work is intended to be compatible with ZFS, the Solaris CIFS
server's interaction with ZFS, somewhat compatible with MacOS X,
and of course compatible with Windows.
The Windows attributes that are implemented were chosen based on
the attributes that ZFS already supports.
The summary of the flags is as follows:
[...]
UF_ARCHIVE:Command line name: uarch, uarchive
ZFS_NAME: XAT_ARCHIVE, ZFS_ARCHIVE
Windows name: FILE_ATTRIBUTE_ARCHIVE
The UF_ARCHIVED flag means that the file has changed and
needs to be archived. The meaning is same as
the Windows FILE_ATTRIBUTE_ARCHIVE attribute, and
the ZFS XAT_ARCHIVE and ZFS_ARCHIVE attribute.
msdosfs and ZFS have special handling for this flag.
i.e. they will set it when the file changes.
Is it intended that this flag is automatically added to all new and
existing ZFS files?
Yes, that is intentional. ZFS already has the flag internally, this just
exposes it to FreeBSD. With ZFS at least, it is set any time a file
changes. If an application clears it on files it has archived, it will
know when the file has changed and it needs to archive the file again.
This is the inverse (more or less) of the SF_ARCHIVED flag.
With UFS, the flag is just passed through and stored.
One application for this is to support CIFS servers that need to store
DOS/CIFS/Windows attributes on a FreeBSD server.
# touch test
# ls -alo test
-rw-r--r--- 1 root wheel uarch 0 Aug 28 15:46 test
This breaks 'cp -p' to tmpfs as tmpfs does not allow this flag.
# mkdir /tmp/tmpfs
# mount -t tmpfs tmpfs /tmp/tmpfs
# cp -f test /tmp/tmpfs
cp: test: Operation not supported
Right. For some filesystems, like UFS and probably tmpfs, the right answer
may be to just pass through most all of the file flags, and reject specific
flags that it doesn't support.
For some filesystems, like smbfs, the right answer seems to be to just
ignore flags that are set that aren't supported.
Other filesystems, like msdosfs, support a small number of flags and reject
any that aren't supported.
In other words, this isn't a new problem, and it would have cropped up if
you tried to copy a file with the SF_IMMUTABLE flag set from UFS to msdosfs
with cp -p.
[...]
Modified: head/sys/ufs/ufs/ufs_vnops.c
==
--- head/sys/ufs/ufs/ufs_vnops.c Wed Aug 21 22:57:29 2013 (r254626)
+++ head/sys/ufs/ufs/ufs_vnops.c Wed Aug 21 23:04:48 2013 (r254627)
@@ -528,9 +528,11 @@ ufs_setattr(ap)
return (EINVAL);
}
if (vap-va_flags != VNOVAL) {
-if ((vap-va_flags ~(UF_NODUMP | UF_IMMUTABLE | UF_APPEND |
-UF_OPAQUE | UF_NOUNLINK | SF_ARCHIVED | SF_IMMUTABLE |
-SF_APPEND | SF_NOUNLINK | SF_SNAPSHOT)) != 0)
+if ((vap-va_flags ~(SF_APPEND | SF_ARCHIVED |
SF_IMMUTABLE |
+SF_NOUNLINK | SF_SNAPSHOT | UF_APPEND | UF_ARCHIVE |
+UF_HIDDEN | UF_IMMUTABLE | UF_NODUMP | UF_NOUNLINK |
+UF_OFFLINE | UF_OPAQUE | UF_READONLY | UF_REPARSE |
+UF_SPARSE | UF_SYSTEM)) != 0)
return (EOPNOTSUPP);
if (vp-v_mount-mnt_flag MNT_RDONLY)
return (EROFS);
Seems a similar change is needed in tmpfs_subr.c:tmpfs_chflags()
(antoine pointed this out)
Sure, I can fix tmpfs.
Ken
--
Kenneth Merry
k...@freebsd.org
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255008 - head/sys/fs/tmpfs
Author: ken Date: Wed Aug 28 22:12:56 2013 New Revision: 255008 URL: http://svnweb.freebsd.org/changeset/base/255008 Log: Support storing 7 additional file flags in tmpfs: UF_SYSTEM, UF_SPARSE, UF_OFFLINE, UF_REPARSE, UF_ARCHIVE, UF_READONLY, and UF_HIDDEN. Sort the file flags tmpfs supports alphabetically. tmpfs now supports the same flags as UFS, with the exception of SF_SNAPSHOT. Reported by: bdrewery, antoine Sponsored by: Spectra Logic Modified: head/sys/fs/tmpfs/tmpfs_subr.c Modified: head/sys/fs/tmpfs/tmpfs_subr.c == --- head/sys/fs/tmpfs/tmpfs_subr.c Wed Aug 28 21:10:37 2013 (r255007) +++ head/sys/fs/tmpfs/tmpfs_subr.c Wed Aug 28 22:12:56 2013 (r255008) @@ -1433,9 +1433,10 @@ tmpfs_chflags(struct vnode *vp, u_long f node = VP_TO_TMPFS_NODE(vp); - if ((flags ~(UF_NODUMP | UF_IMMUTABLE | UF_APPEND | UF_OPAQUE | - UF_NOUNLINK | SF_ARCHIVED | SF_IMMUTABLE | SF_APPEND | - SF_NOUNLINK)) != 0) + if ((flags ~(SF_APPEND | SF_ARCHIVED | SF_IMMUTABLE | SF_NOUNLINK | + UF_APPEND | UF_ARCHIVE | UF_HIDDEN | UF_IMMUTABLE | UF_NODUMP | + UF_NOUNLINK | UF_OFFLINE | UF_OPAQUE | UF_READONLY | UF_REPARSE | + UF_SPARSE | UF_SYSTEM)) != 0) return (EOPNOTSUPP); /* Disallow this operation if the file system is mounted read-only. */ ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
Re: svn commit: r255008 - head/sys/fs/tmpfs
On 2013-08-28 17:12, Kenneth D. Merry wrote: Author: ken Date: Wed Aug 28 22:12:56 2013 New Revision: 255008 URL: http://svnweb.freebsd.org/changeset/base/255008 Log: Support storing 7 additional file flags in tmpfs: UF_SYSTEM, UF_SPARSE, UF_OFFLINE, UF_REPARSE, UF_ARCHIVE, UF_READONLY, and UF_HIDDEN. Sort the file flags tmpfs supports alphabetically. tmpfs now supports the same flags as UFS, with the exception of SF_SNAPSHOT. Reported by: bdrewery, antoine Sponsored by: Spectra Logic Modified: head/sys/fs/tmpfs/tmpfs_subr.c Thanks! ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255009 - head/sys/dev/drm2
Author: jkim
Date: Wed Aug 28 22:57:49 2013
New Revision: 255009
URL: http://svnweb.freebsd.org/changeset/base/255009
Log:
Fix a compiler warning. With this fix, a negative time can be converted to
a struct timeval and back to the original nanoseconds correctly.
Modified:
head/sys/dev/drm2/drm_irq.c
Modified: head/sys/dev/drm2/drm_irq.c
==
--- head/sys/dev/drm2/drm_irq.c Wed Aug 28 22:12:56 2013(r255008)
+++ head/sys/dev/drm2/drm_irq.c Wed Aug 28 22:57:49 2013(r255009)
@@ -210,7 +210,7 @@ struct timeval
ns_to_timeval(const int64_t nsec)
{
struct timeval tv;
- uint32_t rem;
+ long rem;
if (nsec == 0) {
tv.tv_sec = 0;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255010 - head/sys/netinet
Author: np
Date: Wed Aug 28 23:00:34 2013
New Revision: 255010
URL: http://svnweb.freebsd.org/changeset/base/255010
Log:
Merge r254336 from user/np/cxl_tuning.
Add a last-modified timestamp to each LRO entry and provide an interface
to flush all inactive entries. Drivers decide when to flush and what
the inactivity threshold should be.
Network drivers that process an rx queue to completion can enter a
livelock type situation when the rate at which packets are received
reaches equilibrium with the rate at which the rx thread is processing
them. When this happens the final LRO flush (normally when the rx
routine is done) does not occur. Pure ACKs and segments with total
payload 64K can get stuck in an LRO entry. Symptoms are that TCP
tx-mostly connections' performance falls off a cliff during heavy,
unrelated rx on the interface.
Flushing only inactive LRO entries works better than any of these
alternates that I tried:
- don't LRO pure ACKs
- flush _all_ LRO entries periodically (every 'x' microseconds or every
'y' descriptors)
- stop rx processing in the driver periodically and schedule remaining
work for later.
Reviewed by: andre
Modified:
head/sys/netinet/tcp_lro.c
head/sys/netinet/tcp_lro.h
Directory Properties:
head/sys/ (props changed)
Modified: head/sys/netinet/tcp_lro.c
==
--- head/sys/netinet/tcp_lro.c Wed Aug 28 22:57:49 2013(r255009)
+++ head/sys/netinet/tcp_lro.c Wed Aug 28 23:00:34 2013(r255010)
@@ -194,6 +194,25 @@ tcp_lro_rx_csum_fixup(struct lro_entry *
#endif
void
+tcp_lro_flush_inactive(struct lro_ctrl *lc, const struct timeval *timeout)
+{
+ struct lro_entry *le, *le_tmp;
+ struct timeval tv;
+
+ if (SLIST_EMPTY(lc-lro_active))
+ return;
+
+ getmicrotime(tv);
+ timevalsub(tv, timeout);
+ SLIST_FOREACH_SAFE(le, lc-lro_active, next, le_tmp) {
+ if (timevalcmp(tv, le-mtime, =)) {
+ SLIST_REMOVE(lc-lro_active, le, lro_entry, next);
+ tcp_lro_flush(lc, le);
+ }
+ }
+}
+
+void
tcp_lro_flush(struct lro_ctrl *lc, struct lro_entry *le)
{
@@ -543,7 +562,8 @@ tcp_lro_rx(struct lro_ctrl *lc, struct m
if (le-p_len (65535 - lc-ifp-if_mtu)) {
SLIST_REMOVE(lc-lro_active, le, lro_entry, next);
tcp_lro_flush(lc, le);
- }
+ } else
+ getmicrotime(le-mtime);
return (0);
}
@@ -556,6 +576,7 @@ tcp_lro_rx(struct lro_ctrl *lc, struct m
le = SLIST_FIRST(lc-lro_free);
SLIST_REMOVE_HEAD(lc-lro_free, next);
SLIST_INSERT_HEAD(lc-lro_active, le, next);
+ getmicrotime(le-mtime);
/* Start filling in details. */
switch (eh_type) {
Modified: head/sys/netinet/tcp_lro.h
==
--- head/sys/netinet/tcp_lro.h Wed Aug 28 22:57:49 2013(r255009)
+++ head/sys/netinet/tcp_lro.h Wed Aug 28 23:00:34 2013(r255010)
@@ -30,6 +30,8 @@
#ifndef _TCP_LRO_H_
#define _TCP_LRO_H_
+#include sys/time.h
+
struct lro_entry
{
SLIST_ENTRY(lro_entry) next;
@@ -59,6 +61,7 @@ struct lro_entry
uint32_ttsecr;
uint16_twindow;
uint16_ttimestamp; /* flag, not a TCP hdr field. */
+ struct timeval mtime;
};
SLIST_HEAD(lro_head, lro_entry);
@@ -83,6 +86,7 @@ struct lro_ctrl {
int tcp_lro_init(struct lro_ctrl *);
void tcp_lro_free(struct lro_ctrl *);
+void tcp_lro_flush_inactive(struct lro_ctrl *, const struct timeval *);
void tcp_lro_flush(struct lro_ctrl *, struct lro_entry *);
int tcp_lro_rx(struct lro_ctrl *, struct mbuf *, uint32_t);
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255011 - head/sys/dev/cxgbe
Author: np
Date: Wed Aug 28 23:15:05 2013
New Revision: 255011
URL: http://svnweb.freebsd.org/changeset/base/255011
Log:
Whitespace nit.
Modified:
head/sys/dev/cxgbe/t4_tracer.c
Modified: head/sys/dev/cxgbe/t4_tracer.c
==
--- head/sys/dev/cxgbe/t4_tracer.c Wed Aug 28 23:00:34 2013
(r255010)
+++ head/sys/dev/cxgbe/t4_tracer.c Wed Aug 28 23:15:05 2013
(r255011)
@@ -465,7 +465,7 @@ tracer_ioctl(struct ifnet *ifp, unsigned
switch (cmd) {
case SIOCSIFMTU:
case SIOCSIFFLAGS:
- case SIOCADDMULTI:
+ case SIOCADDMULTI:
case SIOCDELMULTI:
case SIOCSIFCAP:
break;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255012 - head/sys/dev/drm2
Author: jkim
Date: Wed Aug 28 23:43:28 2013
New Revision: 255012
URL: http://svnweb.freebsd.org/changeset/base/255012
Log:
Fix a compiler warning and add couple of VM map types.
Modified:
head/sys/dev/drm2/drm_sysctl.c
Modified: head/sys/dev/drm2/drm_sysctl.c
==
--- head/sys/dev/drm2/drm_sysctl.c Wed Aug 28 23:15:05 2013
(r255011)
+++ head/sys/dev/drm2/drm_sysctl.c Wed Aug 28 23:43:28 2013
(r255012)
@@ -177,7 +177,15 @@ static int drm_vm_info DRM_SYSCTL_HANDLE
{
struct drm_device *dev = arg1;
drm_local_map_t *map, *tempmaps;
- const char *types[] = { FB, REG, SHM, AGP, SG };
+ const char *types[] = {
+ [_DRM_FRAME_BUFFER] = FB,
+ [_DRM_REGISTERS] = REG,
+ [_DRM_SHM] = SHM,
+ [_DRM_AGP] = AGP,
+ [_DRM_SCATTER_GATHER] = SG,
+ [_DRM_CONSISTENT] = CONS,
+ [_DRM_GEM] = GEM
+ };
const char *type, *yesno;
int i, mapcount;
char buf[128];
@@ -211,10 +219,20 @@ static int drm_vm_info DRM_SYSCTL_HANDLE
for (i = 0; i mapcount; i++) {
map = tempmaps[i];
- if (map-type 0 || map-type 4)
+ switch(map-type) {
+ default:
type = ??;
- else
+ break;
+ case _DRM_FRAME_BUFFER:
+ case _DRM_REGISTERS:
+ case _DRM_SHM:
+ case _DRM_AGP:
+ case _DRM_SCATTER_GATHER:
+ case _DRM_CONSISTENT:
+ case _DRM_GEM:
type = types[map-type];
+ break;
+ }
if (!map-mtrr)
yesno = no;
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
svn commit: r255013 - head/sys/dev/drm2/i915
Author: jkim
Date: Wed Aug 28 23:59:38 2013
New Revision: 255013
URL: http://svnweb.freebsd.org/changeset/base/255013
Log:
Correct atomic operations in i915.
Modified:
head/sys/dev/drm2/i915/i915_gem.c
head/sys/dev/drm2/i915/i915_gem_execbuffer.c
head/sys/dev/drm2/i915/intel_display.c
Modified: head/sys/dev/drm2/i915/i915_gem.c
==
--- head/sys/dev/drm2/i915/i915_gem.c Wed Aug 28 23:43:28 2013
(r255012)
+++ head/sys/dev/drm2/i915/i915_gem.c Wed Aug 28 23:59:38 2013
(r255013)
@@ -141,7 +141,7 @@ i915_gem_wait_for_error(struct drm_devic
}
mtx_unlock(dev_priv-error_completion_lock);
- if (atomic_read(dev_priv-mm.wedged)) {
+ if (atomic_load_acq_int(dev_priv-mm.wedged)) {
mtx_lock(dev_priv-error_completion_lock);
dev_priv-error_completion++;
mtx_unlock(dev_priv-error_completion_lock);
@@ -743,7 +743,7 @@ i915_gem_ring_throttle(struct drm_device
int ret;
dev_priv = dev-dev_private;
- if (atomic_read(dev_priv-mm.wedged))
+ if (atomic_load_acq_int(dev_priv-mm.wedged))
return (-EIO);
file_priv = file-driver_priv;
@@ -768,15 +768,15 @@ i915_gem_ring_throttle(struct drm_device
if (ring-irq_get(ring)) {
while (ret == 0
!(i915_seqno_passed(ring-get_seqno(ring), seqno) ||
- atomic_read(dev_priv-mm.wedged)))
+ atomic_load_acq_int(dev_priv-mm.wedged)))
ret = -msleep(ring, ring-irq_lock, PCATCH,
915thr, 0);
ring-irq_put(ring);
- if (ret == 0 atomic_read(dev_priv-mm.wedged))
+ if (ret == 0
atomic_load_acq_int(dev_priv-mm.wedged))
ret = -EIO;
} else if (_intel_wait_for(dev,
i915_seqno_passed(ring-get_seqno(ring), seqno) ||
- atomic_read(dev_priv-mm.wedged), 3000, 0, 915rtr)) {
+ atomic_load_acq_int(dev_priv-mm.wedged), 3000, 0,
915rtr)) {
ret = -EBUSY;
}
}
Modified: head/sys/dev/drm2/i915/i915_gem_execbuffer.c
==
--- head/sys/dev/drm2/i915/i915_gem_execbuffer.cWed Aug 28 23:43:28
2013(r255012)
+++ head/sys/dev/drm2/i915/i915_gem_execbuffer.cWed Aug 28 23:59:38
2013(r255013)
@@ -192,7 +192,7 @@ i915_gem_object_set_to_gpu_domain(struct
i915_gem_clflush_object(obj);
if (obj-base.pending_write_domain)
- cd-flips |= atomic_read(obj-pending_flip);
+ cd-flips |= atomic_load_acq_int(obj-pending_flip);
/* The actual obj-write_domain will be updated with
* pending_write_domain after we emit the accumulated flush for all
Modified: head/sys/dev/drm2/i915/intel_display.c
==
--- head/sys/dev/drm2/i915/intel_display.c Wed Aug 28 23:43:28 2013
(r255012)
+++ head/sys/dev/drm2/i915/intel_display.c Wed Aug 28 23:59:38 2013
(r255013)
@@ -2261,8 +2261,8 @@ intel_finish_fb(struct drm_framebuffer *
int ret;
mtx_lock(dev-event_lock);
- while (!atomic_read(dev_priv-mm.wedged)
- atomic_read(obj-pending_flip) != 0) {
+ while (!atomic_load_acq_int(dev_priv-mm.wedged)
+ atomic_load_acq_int(obj-pending_flip) != 0) {
msleep(obj-pending_flip, dev-event_lock,
0, 915flp, 0);
}
@@ -2948,7 +2948,7 @@ static void intel_crtc_wait_for_pending_
dev = crtc-dev;
dev_priv = dev-dev_private;
mtx_lock(dev-event_lock);
- while (atomic_read(obj-pending_flip) != 0)
+ while (atomic_load_acq_int(obj-pending_flip) != 0)
msleep(obj-pending_flip, dev-event_lock, 0, 915wfl, 0);
mtx_unlock(dev-event_lock);
}
@@ -7333,7 +7333,7 @@ static void do_intel_finish_page_flip(st
obj = work-old_fb_obj;
atomic_clear_int(obj-pending_flip, 1 intel_crtc-plane);
- if (atomic_read(obj-pending_flip) == 0)
+ if (atomic_load_acq_int(obj-pending_flip) == 0)
wakeup(obj-pending_flip);
mtx_unlock(dev-event_lock);
@@ -7640,7 +7640,7 @@ static int intel_crtc_page_flip(struct d
return 0;
cleanup_pending:
- atomic_sub(1 intel_crtc-plane, work-old_fb_obj-pending_flip);
+ atomic_clear_int(work-old_fb_obj-pending_flip, 1
intel_crtc-plane);
drm_gem_object_unreference(work-old_fb_obj-base);
drm_gem_object_unreference(obj-base);
DRM_UNLOCK(dev);
___
svn-src-all@freebsd.org mailing list
svn commit: r255014 - in head: gnu/usr.bin/patch share/mk tools/build/options usr.bin/patch
Author: pfg
Date: Thu Aug 29 00:38:24 2013
New Revision: 255014
URL: http://svnweb.freebsd.org/changeset/base/255014
Log:
Drop build option switch for the older GNU patch.
As promised, drop the option to make the older GNU patch
the default.
GNU patch is still being built but something drastic may
happen to it to it before Release.
Deleted:
head/tools/build/options/WITH_GNU_PATCH
Modified:
head/gnu/usr.bin/patch/Makefile
head/share/mk/bsd.own.mk
head/usr.bin/patch/Makefile
Modified: head/gnu/usr.bin/patch/Makefile
==
--- head/gnu/usr.bin/patch/Makefile Wed Aug 28 23:59:38 2013
(r255013)
+++ head/gnu/usr.bin/patch/Makefile Thu Aug 29 00:38:24 2013
(r255014)
@@ -1,16 +1,10 @@
# $FreeBSD$
-.include bsd.own.mk
-
-.if ${MK_GNU_PATCH} == yes
-PROG= patch
-.else
PROG= gnupatch
CLEANFILES+= gnupatch.1
gnupatch.1: patch.1
cp ${.ALLSRC} ${.TARGET}
-.endif
SRCS= backupfile.c inp.c patch.c pch.c util.c version.c
CFLAGS+=-DHAVE_CONFIG_H
Modified: head/share/mk/bsd.own.mk
==
--- head/share/mk/bsd.own.mkWed Aug 28 23:59:38 2013(r255013)
+++ head/share/mk/bsd.own.mkThu Aug 29 00:38:24 2013(r255014)
@@ -373,7 +373,6 @@ __DEFAULT_NO_OPTIONS = \
CLANG_EXTRAS \
CTF \
DEBUG_FILES \
-GNU_PATCH \
GPL_DTC \
HESIOD \
LIBICONV_COMPAT \
Modified: head/usr.bin/patch/Makefile
==
--- head/usr.bin/patch/Makefile Wed Aug 28 23:59:38 2013(r255013)
+++ head/usr.bin/patch/Makefile Thu Aug 29 00:38:24 2013(r255014)
@@ -1,17 +1,7 @@
# $OpenBSD: Makefile,v 1.4 2005/05/16 15:22:46 espie Exp $
# $FreeBSD$
-.include bsd.own.mk
-
-.if ${MK_GNU_PATCH} == yes
-PROG= bsdpatch
-CLEANFILES+= bsdpatch.1
-
-bsdpatch.1: patch.1
- cp ${.ALLSRC} ${.TARGET}
-.else
PROG= patch
-.endif
SRCS= backupfile.c inp.c mkpath.c patch.c pch.c util.c
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org
