Re: svn commit: r368736 - head/tools/tools/git/hooks
On Mon, Dec 28, 2020 at 3:29 PM Ed Maste wrote: > On Mon, 28 Dec 2020 at 14:34, Alan Somers wrote: > > > > On Thu, Dec 17, 2020 at 12:58 PM Ed Maste wrote: > >> > >> Author: emaste > >> Date: Thu Dec 17 19:58:29 2020 > >> New Revision: 368736 > >> URL: https://svnweb.freebsd.org/changeset/base/368736 > >> > >> Log: > >> Add initial version of git commit message preparation hook > >> > > Thanks! But how can we use it? Is there a "git config" incantation > that will use this file? > > Just copy it into .git/hooks/. Can also fetch the latest version > directly from cgit; lwhsu added this to > https://github.com/bsdimp/freebsd-git-docs/blob/main/URLs.md: > > fetch > https://cgit.freebsd.org/src/plain/tools/tools/git/hooks/prepare-commit-msg > -o > <https://cgit.freebsd.org/src/plain/tools/tools/git/hooks/prepare-commit-msg-o> > .git/hooks > chmod 755 .git/hooks/prepare-commit-msg > > I'd like us to end up with a setup-repo.sh or such that will offer to > take care of all of these things. > > >Also, this seems too complicated. Git already has a "config.template" > configuration variable. Could we just use that instead, or does it not > work for some reason? > > It does work, although it's a bit awkward. For one thing, the provided > template is placed at the top of the message, followed by the default > message, but our addition is really commented-out information and fits > best in the middle of the existing text - after the "Please enter > the..." and before the list of changes and such. With config.template > we'd end up with something like: > > > > # PR: > # ... > # Differential Revision:<https://reviews.freebsd.org/D###> > # > # Please enter the commit message for your changes. Lines starting > # with '#' will be ignored, and an empty message aborts the commit. > # > # ... > > > Additionally we might want to make some more programmatic changes in > the default message (depending on the resolution of some open > questions). For example we could strip the "MFC after" line that's > copied from the original commit, when preparing the commit message for > a MFC/cherry-pick. > Thanks for the explanation. It all makes sense, and the commit template works. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r366697 - head/usr.bin/xinstall
BTW, I have a WIP patch to xinstall to make it use copy_file_range. The
patch works, but I never wrote a fallback path in case copy_file_range
fails for some reason. Alex, would you be interested to finish the patch?
-Alan
On Wed, Oct 14, 2020 at 7:35 AM Alexander Richardson <
arichard...@freebsd.org> wrote:
> On Wed, 14 Oct 2020 at 14:29, Mateusz Guzik wrote:
> >
> > This should use copy_file_range (also available on Linux).
> >
>
> I agree. I even mentioned this in
> https://reviews.freebsd.org/D26041#589287.
> This change avoids the two unnecessary syscalls, but I agree that
> longer-term install should share the copy_file_range code with cp.
> The only thing that copy_file_range won't speed up is the check
> whether source and target are already identical.
>
> Alex
>
> > On 10/14/20, Alex Richardson wrote:
> > > Author: arichardson
> > > Date: Wed Oct 14 12:28:41 2020
> > > New Revision: 366697
> > > URL: https://svnweb.freebsd.org/changeset/base/366697
> > >
> > > Log:
> > > install(1): Avoid unncessary fstatfs() calls and use mmap() based on
> size
> > >
> > > According to git blame the trymmap() function was added in 1996 to
> skip
> > > mmap() calls for NFS file systems. However, nowadays mmap() should be
> > > perfectly safe even on NFS. Importantly, onl ufs and cd9660 file
> systems
> > > were whitelisted so we don't use mmap() on ZFS. It also prevents the
> use
> > > of mmap() when bootstrapping from macOS/Linux since on those systems
> the
> > > trymmap() function was always returning zero due to the missing
> > > MFSNAMELEN
> > > define.
> > >
> > > This change keeps the trymmap() function but changes it to check
> whether
> > > using mmap() can reduce the number of system calls that are required.
> > > Using mmap() only reduces the number of system calls if we need
> multiple
> > > read()
> > > syscalls, i.e. if the file size is > MAXBSIZE. However, mmap() is
> more
> > > expensive
> > > than read() so this sets the threshold at 4 fewer syscalls.
> Additionally,
> > > for
> > > larger file size mmap() can significantly increase the number of page
> > > faults,
> > > so avoid it in that case.
> > >
> > > It's unclear whether using mmap() is ever faster than a read with an
> > > appropriate
> > > buffer size, but this change at least removes two unnecessary system
> > > calls
> > > for every file that is installed.
> > >
> > > Reviewed By:markj
> > > Differential Revision: https://reviews.freebsd.org/D26041
> > >
> > > Modified:
> > > head/usr.bin/xinstall/xinstall.c
> > >
> > > Modified: head/usr.bin/xinstall/xinstall.c
> > >
> ==
> > > --- head/usr.bin/xinstall/xinstall.c Wed Oct 14 10:12:39 2020
> (r366696)
> > > +++ head/usr.bin/xinstall/xinstall.c Wed Oct 14 12:28:41 2020
> (r366697)
> > > @@ -148,7 +148,7 @@ static void metadata_log(const char *, const
> char *, s
> > > const char *, const char *, off_t);
> > > static int parseid(const char *, id_t *);
> > > static int strip(const char *, int, const char *, char **);
> > > -static int trymmap(int);
> > > +static int trymmap(size_t);
> > > static void usage(void);
> > >
> > > int
> > > @@ -1087,7 +1087,7 @@ compare(int from_fd, const char *from_name
> __unused,
> > > s
> > > if (do_digest)
> > > digest_init();
> > > done_compare = 0;
> > > - if (trymmap(from_fd) && trymmap(to_fd)) {
> > > + if (trymmap(from_len) && trymmap(to_len)) {
> > > p = mmap(NULL, from_len, PROT_READ, MAP_SHARED,
> > > from_fd, (off_t)0);
> > > if (p == MAP_FAILED)
> > > @@ -1248,13 +1248,8 @@ copy(int from_fd, const char *from_name, int
> to_fd,
> > > co
> > >
> > > digest_init();
> > >
> > > - /*
> > > - * Mmap and write if less than 8M (the limit is so we don't
> totally
> > > - * trash memory on big files. This is really a minor hack, but
> it
> > > - * wins some CPU back.
> > > - */
> > > done_copy = 0;
> > > - if (size <= 8 * 1048576 && trymmap(from_fd) &&
> > > + if (trymmap((size_t)size) &&
> > > (p = mmap(NULL, (size_t)size, PROT_READ, MAP_SHARED,
> > > from_fd, (off_t)0)) != MAP_FAILED) {
> > > nw = write(to_fd, p, size);
> > > @@ -1523,20 +1518,23 @@ usage(void)
> > > * return true (1) if mmap should be tried, false (0) if not.
> > > */
> > > static int
> > > -trymmap(int fd)
> > > +trymmap(size_t filesize)
> > > {
> > > -/*
> > > - * The ifdef is for bootstrapping - f_fstypename doesn't exist in
> > > - * pre-Lite2-merge systems.
> > > - */
> > > -#ifdef MFSNAMELEN
> > > - struct statfs stfs;
> > > -
> > > - if (fstatfs(fd, ) != 0)
> > > - return (0);
> > > - if (strcmp(stfs.f_fstypename, "ufs") == 0 ||
> > > -
Re: svn commit: r368736 - head/tools/tools/git/hooks
On Thu, Dec 17, 2020 at 12:58 PM Ed Maste wrote: > Author: emaste > Date: Thu Dec 17 19:58:29 2020 > New Revision: 368736 > URL: https://svnweb.freebsd.org/changeset/base/368736 > > Log: > Add initial version of git commit message preparation hook > > Start with a slightly modified version of the SVN commit template, to > allow developers to experiment. This will be updated in the future as > our process and techniques evolve. > > This can be installed by copying or symlinking into the .git/hooks/ > directory. > > Feedback from:cem, jhb > Reviewed by: imp > Sponsored by: The FreeBSD Foundation > Differential Revision:https://reviews.freebsd.org/D27633 > > Added: > head/tools/tools/git/hooks/ > head/tools/tools/git/hooks/prepare-commit-msg (contents, props changed) > Thanks! But how can we use it? Is there a "git config" incantation that will use this file? Also, this seems too complicated. Git already has a "config.template" configuration variable. Could we just use that instead, or does it not work for some reason? -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365917 - in releng/12.2: cddl/contrib/opensolaris/cmd/zfs cddl/contrib/opensolaris/cmd/zpool sys/cddl/compat/opensolaris/kern sys/cddl/compat/opensolaris/sys sys/cddl/contrib/opensola
On Sat, Sep 19, 2020 at 2:47 PM Allan Jude wrote: > Author: allanjude > Date: Sat Sep 19 20:46:56 2020 > New Revision: 365917 > URL: https://svnweb.freebsd.org/changeset/base/365917 > > Log: > MFS r365689,r365808,r365860 > > MFOpenZFS: Introduce read/write kstats per dataset > > The following patch introduces a few statistics on reads and writes > grouped by dataset. These statistics are implemented as kstats > (backed by aggregate sums for performance) and can be retrieved by > using the dataset objset ID number. The motivation for this change is > to provide some preliminary analytics on dataset usage/performance. > > Reviewed-by: Richard Elling > Reviewed-by: Brian Behlendorf > Reviewed by: Matthew Ahrens > Signed-off-by: Serapheim Dimitropoulos > > openzfs/zfs@a448a2557ec4938ed6944c7766fe0b8e6e5f6456 > > Also contains parts of: > MFOpenZFS: Connect dataset_kstats for FreeBSD > > Example output: > kstat.zfs.mypool.dataset.objset-0x10b.nread: 150528 > kstat.zfs.mypool.dataset.objset-0x10b.reads: 48 > kstat.zfs.mypool.dataset.objset-0x10b.nwritten: 134217728 > kstat.zfs.mypool.dataset.objset-0x10b.writes: 1024 > kstat.zfs.mypool.dataset.objset-0x10b.dataset_name: mypool/datasetname > > Reviewed-by: Ryan Moeller > Reviewed by: Sean Eric Fagan > Reviewed-by: Serapheim Dimitropoulos > Reviewed-by: Brian Behlendorf > Signed-off-by: Allan Jude > > openzfs/zfs@4547fc4e071ceb1818b3a46c3035b923e06e5390 > > Approved by: re (gjb) > Relnotes: yes > Sponsored by: Klara Inc. > This sounds quite useful! I've often wished that I could see per-dataset traffic statistics. Are there any user-friendly frontends to this? Like zfs-stats, or an updated zpool iostat? -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368597 - stable/12/sbin/ping6
Author: asomers Date: Sat Dec 12 23:01:11 2020 New Revision: 368597 URL: https://svnweb.freebsd.org/changeset/base/368597 Log: MFC r367976: ping6: update usage text after r365547 Modified: stable/12/sbin/ping6/ping6.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sbin/ping6/ping6.c == --- stable/12/sbin/ping6/ping6.cSat Dec 12 22:57:28 2020 (r368596) +++ stable/12/sbin/ping6/ping6.cSat Dec 12 23:01:11 2020 (r368597) @@ -2838,7 +2838,7 @@ usage(void) "\n" " [-p pattern] [-S sourceaddr] [-s packetsize] " "[-x waittime]\n" - " [-X timeout] [hops ...] host\n"); + " [-X timeout] [-z tclass] [hops ...] host\n"); exit(1); } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368596 - stable/12/tests/sys/fs/fusefs
Author: asomers
Date: Sat Dec 12 22:57:28 2020
New Revision: 368596
URL: https://svnweb.freebsd.org/changeset/base/368596
Log:
MFC r366365:
fusefs tests: quell Coverity "Argument cannot be negative" warnings
Must abort tests early if open(2) fails.
Reported by: Coverity
Coverity CID: 1432810 and many others
Reviewed by: kevans
Differential Revision:https://reviews.freebsd.org/D26635
Modified:
stable/12/tests/sys/fs/fusefs/allow_other.cc
stable/12/tests/sys/fs/fusefs/create.cc
stable/12/tests/sys/fs/fusefs/default_permissions.cc
stable/12/tests/sys/fs/fusefs/flush.cc
stable/12/tests/sys/fs/fusefs/open.cc
stable/12/tests/sys/fs/fusefs/opendir.cc
stable/12/tests/sys/fs/fusefs/release.cc
stable/12/tests/sys/fs/fusefs/releasedir.cc
stable/12/tests/sys/fs/fusefs/write.cc
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/tests/sys/fs/fusefs/allow_other.cc
==
--- stable/12/tests/sys/fs/fusefs/allow_other.ccSat Dec 12 22:51:30
2020(r368595)
+++ stable/12/tests/sys/fs/fusefs/allow_other.ccSat Dec 12 22:57:28
2020(r368596)
@@ -168,7 +168,7 @@ TEST_F(AllowOther, privilege_escalation)
.WillRepeatedly(Invoke(ReturnErrno(EPERM)));
fd1 = open(FULLPATH, O_RDONLY);
- EXPECT_LE(0, fd1) << strerror(errno);
+ ASSERT_LE(0, fd1) << strerror(errno);
}, [] {
int fd0;
Modified: stable/12/tests/sys/fs/fusefs/create.cc
==
--- stable/12/tests/sys/fs/fusefs/create.cc Sat Dec 12 22:51:30 2020
(r368595)
+++ stable/12/tests/sys/fs/fusefs/create.cc Sat Dec 12 22:57:28 2020
(r368596)
@@ -143,7 +143,7 @@ TEST_F(Create, attr_cache)
).Times(0);
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -184,7 +184,7 @@ TEST_F(Create, clear_attr_cache)
EXPECT_EQ(0, stat("mountpoint", )) << strerror(errno);
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
EXPECT_EQ(0, stat("mountpoint", )) << strerror(errno);
leak(fd);
@@ -254,7 +254,7 @@ TEST_F(Create, Enosys)
})));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -366,7 +366,7 @@ TEST_F(Create, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -398,7 +398,7 @@ TEST_F(Create, wronly_0444)
}));
fd = open(FULLPATH, O_CREAT | O_WRONLY, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -422,7 +422,7 @@ TEST_F(Create_7_8, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -446,6 +446,6 @@ TEST_F(Create_7_11, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
Modified: stable/12/tests/sys/fs/fusefs/default_permissions.cc
==
--- stable/12/tests/sys/fs/fusefs/default_permissions.ccSat Dec 12
22:51:30 2020(r368595)
+++ stable/12/tests/sys/fs/fusefs/default_permissions.ccSat Dec 12
22:57:28 2020(r368596)
@@ -490,7 +490,7 @@ TEST_F(Create, ok)
expect_create(RELPATH, ino);
fd = open(FULLPATH, O_CREAT | O_EXCL, 0644);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -765,7 +765,7 @@ TEST_F(Open, ok)
expect_open(ino, 0, 1);
fd = open(FULLPATH, O_RDONLY);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
Modified: stable/12/tests/sys/fs/fusefs/flush.cc
==
--- stable/12/tests/sys/fs/fusefs/flush.cc Sat Dec 12 22:51:30 2020
(r368595)
+++ stable/12/tests/sys/fs/fusefs/flush.cc Sat Dec 12 22:57:28 2020
(r368596)
@@ -102,10 +102,10 @@ TEST_F(Flush, open_twice)
expect_release();
fd = open(FULLPATH, O_WRONLY);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
fd2 = open(FULLPATH, O_WRONLY);
- EXPECT_LE(0, fd2) << strerror(errno);
+
svn commit: r368491 - stable/12/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Wed Dec 9 20:06:37 2020
New Revision: 368491
URL: https://svnweb.freebsd.org/changeset/base/368491
Log:
ZFS: fix spurious EBUSY after zfs receive to an existing dataset
If you do a "zfs send -p | zfs receive -F " to an existing but
empty dataset, the receive will complete successfully but spuriously fail
with exit status 1 and the message "cannot mount 'pool/dataset': mountpoint
or dataset is busy".
The root cause is a merge error made in r344569 and MFCed in r345578, which
merged changes a10d50f999 and e63ac16d25 from ZoL. The merge:
* failed to flip a == to an != like the upstream change did, and
* Left out one chunk
Direct commit to stable/12 because head has moved on to OpenZFS.
PR: 251694
Reviewed by: bapt
Sponsored by: Axcient
Modified:
stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c
Modified: stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c
==
--- stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c
Wed Dec 9 18:43:58 2020(r368490)
+++ stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c
Wed Dec 9 20:06:37 2020(r368491)
@@ -895,13 +895,25 @@ libzfs_mnttab_add(libzfs_handle_t *hdl, const char *sp
mnttab_node_t *mtn;
pthread_mutex_lock(>libzfs_mnttab_cache_lock);
- if (avl_numnodes(>libzfs_mnttab_cache) == 0) {
+ if (avl_numnodes(>libzfs_mnttab_cache) != 0) {
mtn = zfs_alloc(hdl, sizeof (mnttab_node_t));
mtn->mtn_mt.mnt_special = zfs_strdup(hdl, special);
mtn->mtn_mt.mnt_mountp = zfs_strdup(hdl, mountp);
mtn->mtn_mt.mnt_fstype = zfs_strdup(hdl, MNTTYPE_ZFS);
mtn->mtn_mt.mnt_mntopts = zfs_strdup(hdl, mntopts);
- avl_add(>libzfs_mnttab_cache, mtn);
+ /*
+* Another thread may have already added this entry
+* via libzfs_mnttab_update. If so we should skip it.
+*/
+ if (avl_find(>libzfs_mnttab_cache, mtn, NULL) != NULL) {
+ free(mtn->mtn_mt.mnt_special);
+ free(mtn->mtn_mt.mnt_mountp);
+ free(mtn->mtn_mt.mnt_fstype);
+ free(mtn->mtn_mt.mnt_mntopts);
+ free(mtn);
+ } else {
+ avl_add(>libzfs_mnttab_cache, mtn);
+ }
}
pthread_mutex_unlock(>libzfs_mnttab_cache_lock);
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368272 - head/tests/sys/aio
Author: asomers
Date: Wed Dec 2 00:53:03 2020
New Revision: 368272
URL: https://svnweb.freebsd.org/changeset/base/368272
Log:
AIO tests: update expected failure messages after r368265
PR: 220398, 251515
MFC after:1 week
MFC-With: r368265
Modified:
head/tests/sys/aio/lio_test.c
Modified: head/tests/sys/aio/lio_test.c
==
--- head/tests/sys/aio/lio_test.c Wed Dec 2 00:48:15 2020
(r368271)
+++ head/tests/sys/aio/lio_test.c Wed Dec 2 00:53:03 2020
(r368272)
@@ -143,8 +143,8 @@ ATF_TC_BODY(lio_listio_empty_nowait_kevent, tc)
int kq, result;
void *udata = (void*)0xdeadbeefdeadbeef;
- atf_tc_expect_timeout("Bug 220398 - lio_listio(2) never sends"
- " asynchronous notification if nent==0");
+ atf_tc_expect_timeout("Bug 251515 - lio_listio(2) never sends"
+ " kevent if nent==0");
kq = kqueue();
ATF_REQUIRE(kq > 0);
sev.sigev_notify = SIGEV_KEVENT;
@@ -168,8 +168,6 @@ ATF_TC_BODY(lio_listio_empty_nowait_signal, tc)
struct aiocb *list = NULL;
struct sigevent sev;
- atf_tc_expect_timeout("Bug 220398 - lio_listio(2) never sends"
- " asynchronous notification if nent==0");
ATF_REQUIRE_EQ(0, sem_init(, false, 0));
sev.sigev_notify = SIGEV_SIGNAL;
sev.sigev_signo = SIGUSR1;
@@ -189,9 +187,6 @@ ATF_TC_BODY(lio_listio_empty_nowait_thread, tc)
struct aiocb *list = NULL;
struct sigevent sev;
- atf_tc_skip("Sometimes hangs and sometimes passes");
- atf_tc_expect_timeout("Bug 220398 - lio_listio(2) never sends"
- " asynchronous notification if nent==0");
ATF_REQUIRE_EQ(0, sem_init(, false, 0));
bzero(, sizeof(sev));
sev.sigev_notify = SIGEV_THREAD;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368233 - in stable/12: cddl/contrib/opensolaris/lib/libzfs/common sys/cddl/contrib/opensolaris/uts/common/sys/fs
Author: asomers
Date: Tue Dec 1 15:15:18 2020
New Revision: 368233
URL: https://svnweb.freebsd.org/changeset/base/368233
Log:
Fix error merging r354116 from OpenZFS
When we merged 4c0883fb4af0d5565459099b98fcf90ecbfa1ca1 from OpenZFS (svn
r354116), there were some merge conflicts. One of those was resolved
incorrectly, causing "zfs receive" to fail to delete snapshots that a "zfs
send -R" stream has deleted.
This change corrects that merge conflict, and also reduces some harmless
diffs vis-a-vis OpenZFS that were also introduced by the same revision.
Direct commit to stable/12 because head has moved on to OpenZFS.
PR: 249438
Reported by: Dmitry Wagin
Reviewed by: mmacy
Sponsored by: Axcient
Modified:
stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
stable/12/sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h
Modified: stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Tue Dec 1 15:11:16 2020(r368232)
+++ stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Tue Dec 1 15:15:18 2020(r368233)
@@ -613,8 +613,8 @@ typedef struct send_data {
const char *fromsnap;
const char *tosnap;
boolean_t recursive;
- boolean_t verbose;
boolean_t replicate;
+ boolean_t verbose;
/*
* The header nvlist is of the following format:
@@ -848,36 +848,36 @@ send_iterate_fs(zfs_handle_t *zhp, void *arg)
rv = -1;
goto out;
}
- VERIFY(0 == nvlist_add_uint64(nvfs, "origin",
- origin->zfs_dmustats.dds_guid));
+ fnvlist_add_uint64(nvfs, "origin",
+ origin->zfs_dmustats.dds_guid);
}
/* iterate over props */
- VERIFY(0 == nvlist_alloc(, NV_UNIQUE_NAME, 0));
+ nv = fnvlist_alloc();
send_iterate_prop(zhp, nv);
- VERIFY(0 == nvlist_add_nvlist(nvfs, "props", nv));
- nvlist_free(nv);
+ fnvlist_add_nvlist(nvfs, "props", nv);
+ fnvlist_free(nv);
/* iterate over snaps, and set sd->parent_fromsnap_guid */
+ sd->parent_fromsnap_guid = 0;
+ sd->parent_snaps = fnvlist_alloc();
+ sd->snapprops = fnvlist_alloc();
if (!sd->replicate && fromsnap_txg != 0)
min_txg = fromsnap_txg;
if (!sd->replicate && tosnap_txg != 0)
max_txg = tosnap_txg;
- sd->parent_fromsnap_guid = 0;
- VERIFY(0 == nvlist_alloc(>parent_snaps, NV_UNIQUE_NAME, 0));
- VERIFY(0 == nvlist_alloc(>snapprops, NV_UNIQUE_NAME, 0));
(void) zfs_iter_snapshots_sorted(zhp, send_iterate_snap, sd,
min_txg, max_txg);
- VERIFY(0 == nvlist_add_nvlist(nvfs, "snaps", sd->parent_snaps));
- VERIFY(0 == nvlist_add_nvlist(nvfs, "snapprops", sd->snapprops));
+ fnvlist_add_nvlist(nvfs, "snaps", sd->parent_snaps);
+ fnvlist_add_nvlist(nvfs, "snapprops", sd->snapprops);
fnvlist_free(sd->parent_snaps);
fnvlist_free(sd->snapprops);
/* add this fs to nvlist */
(void) snprintf(guidstring, sizeof (guidstring),
"0x%llx", (longlong_t)guid);
- VERIFY(0 == nvlist_add_nvlist(sd->fss, guidstring, nvfs));
- nvlist_free(nvfs);
+ fnvlist_add_nvlist(sd->fss, guidstring, nvfs);
+ fnvlist_free(nvfs);
/* iterate over children */
if (sd->recursive)
@@ -894,13 +894,12 @@ out:
static int
gather_nvlist(libzfs_handle_t *hdl, const char *fsname, const char *fromsnap,
-const char *tosnap, boolean_t recursive, boolean_t verbose,
-boolean_t replicate, nvlist_t **nvlp, avl_tree_t **avlp)
+const char *tosnap, boolean_t recursive, boolean_t replicate,
+boolean_t verbose, nvlist_t **nvlp, avl_tree_t **avlp)
{
zfs_handle_t *zhp;
- int error;
- uint64_t min_txg = 0, max_txg = 0;
send_data_t sd = { 0 };
+ int error;
zhp = zfs_open(hdl, fsname, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME);
if (zhp == NULL)
@@ -911,8 +910,8 @@ gather_nvlist(libzfs_handle_t *hdl, const char *fsname
sd.fromsnap = fromsnap;
sd.tosnap = tosnap;
sd.recursive = recursive;
- sd.verbose = verbose;
sd.replicate = replicate;
+ sd.verbose = verbose;
if ((error = send_iterate_fs(zhp, )) != 0) {
nvlist_free(sd.fss);
@@ -1349,10 +1348,10 @@ static int
dump_filesystem(zfs_handle_t *zhp, void *arg)
{
int rv = 0;
- uint64_t min_txg = 0, max_txg = 0;
send_dump_data_t *sdd = arg;
boolean_t missingfrom = B_FALSE;
zfs_cmd_t zc = { 0 };
+ uint64_t min_txg = 0, max_txg = 0;
(void) snprintf(zc.zc_name, sizeof (zc.zc_name), "%s@%s",
svn commit: r368131 - head/sbin/ping
Author: asomers
Date: Sat Nov 28 23:24:19 2020
New Revision: 368131
URL: https://svnweb.freebsd.org/changeset/base/368131
Log:
ping: allow building without INET support
Building without INET6 support was already possible. Now it's possible to
build ping with only INET6, or even with neither INET nor INET6.
Reported by: bz
Reviewed by: bz
MFC-With: 368045
Differential Revision:https://reviews.freebsd.org/D27394
Modified:
head/sbin/ping/Makefile
head/sbin/ping/main.c
Modified: head/sbin/ping/Makefile
==
--- head/sbin/ping/Makefile Sat Nov 28 22:34:33 2020(r368130)
+++ head/sbin/ping/Makefile Sat Nov 28 23:24:19 2020(r368131)
@@ -5,11 +5,16 @@
PACKAGE=runtime
PROG= ping
-SRCS= main.c ping.c utils.c
+SRCS= main.c
MAN= ping.8
BINOWN=root
BINMODE=4555
LIBADD=m
+
+.if ${MK_INET_SUPPORT}!= "no"
+CFLAGS+= -DINET
+SRCS+= ping.c utils.c
+.endif
.if ${MK_INET6_SUPPORT} != "no"
CFLAGS+= -DINET6 -DKAME_SCOPEID
Modified: head/sbin/ping/main.c
==
--- head/sbin/ping/main.c Sat Nov 28 22:34:33 2020(r368130)
+++ head/sbin/ping/main.c Sat Nov 28 23:24:19 2020(r368131)
@@ -44,40 +44,51 @@ __FBSDID("$FreeBSD$");
#include
#include "main.h"
+#ifdef INET
#include "ping.h"
+#endif
#ifdef INET6
#include "ping6.h"
#endif
-#ifdef INET6
+#if defined(INET) && defined(INET6)
#defineOPTSTR ":46"
-#else
+#elif defined(INET)
#define OPTSTR ":4"
+#elif defined(INET6)
+#defineOPTSTR ":6"
+#else
+#define OPTSTR ""
#endif
int
main(int argc, char *argv[])
{
+#if defined(INET) && defined(INET6)
struct in_addr a;
+ struct in6_addr a6;
+#endif
+#if defined(INET) || defined(INET6)
struct addrinfo hints;
+#endif
int ch;
- bool ipv4;
+#ifdef INET
+ bool ipv4 = false;
+#endif
#ifdef INET6
- struct in6_addr a6;
- bool ipv6;
+ bool ipv6 = false;
if (strcmp(getprogname(), "ping6") == 0)
ipv6 = true;
- else
- ipv6 = false;
#endif
- ipv4 = false;
while ((ch = getopt(argc, argv, OPTSTR)) != -1) {
switch(ch) {
+#ifdef INET
case '4':
ipv4 = true;
break;
+#endif
#ifdef INET6
case '6':
ipv6 = true;
@@ -93,20 +104,18 @@ main(int argc, char *argv[])
optreset = 1;
optind = 1;
-#ifdef INET6
+#if defined(INET) && defined(INET6)
if (ipv4 && ipv6)
errx(1, "-4 and -6 cannot be used simultaneously");
#endif
+#if defined(INET) && defined(INET6)
if (inet_pton(AF_INET, argv[argc - 1], ) == 1) {
-#ifdef INET6
if (ipv6)
errx(1, "IPv6 requested but IPv4 target address "
"provided");
-#endif
hints.ai_family = AF_INET;
}
-#ifdef INET6
else if (inet_pton(AF_INET6, argv[argc - 1], ) == 1) {
if (ipv4)
errx(1, "IPv4 requested but IPv6 target address "
@@ -114,7 +123,6 @@ main(int argc, char *argv[])
hints.ai_family = AF_INET6;
} else if (ipv6)
hints.ai_family = AF_INET6;
-#endif
else if (ipv4)
hints.ai_family = AF_INET;
else {
@@ -129,22 +137,30 @@ main(int argc, char *argv[])
freeaddrinfo(res);
}
}
+#elif defined(INET)
+ hints.ai_family = AF_INET;
+#elif defined(INET6)
+ hints.ai_family = AF_INET6;
+#endif
+#ifdef INET
if (hints.ai_family == AF_INET)
return ping(argc, argv);
+#endif /* INET */
#ifdef INET6
- else if (hints.ai_family == AF_INET6)
+ if (hints.ai_family == AF_INET6)
return ping6(argc, argv);
-#endif
- else
- errx(1, "Unknown host");
+#endif /* INET6 */
+ errx(1, "Unknown host");
}
void
usage(void)
{
(void)fprintf(stderr,
- "usage: ping [-4AaDdfHnoQqRrv] [-C pcp] [-c count] "
+ "usage:\n"
+#ifdef INET
+ "\tping [-4AaDdfHnoQqRrv] [-C pcp] [-c count] "
"[-G sweepmaxsize]\n"
" [-g sweepminsize] [-h sweepincrsize] [-i wait] "
"[-l preload]\n"
@@ -155,7 +171,7 @@ usage(void)
"[-p pattern] [-S src_addr] \n"
" [-s packetsize] [-t timeout] [-W waittime] [-z tos] "
"IPv4-host\n"
- " ping [-4AaDdfHLnoQqRrv] [-C pcp] [-c count] [-I iface] "
+ "\tping [-4AaDdfHLnoQqRrv] [-C pcp] [-c count] [-I iface] "
"[-i wait]\n"
" [-l preload] [-M mask | time] [-m ttl] "
#ifdef IPSEC
@@ -164,8 +180,9 @@ usage(void)
"[-p pattern]\n"
"
Re: svn commit: r368045 - in head: . etc/mtree rescue/rescue sbin sbin/ping sbin/ping/tests sbin/ping6 sbin/ping6/tests tools/build/mk
On Fri, Nov 27, 2020 at 5:08 AM Bjoern A. Zeeb < bzeeb-li...@lists.zabbadoz.net> wrote: > On 26 Nov 2020, at 18:51, Alan Somers wrote: > > > On Thu, Nov 26, 2020 at 2:16 AM Bjoern A. Zeeb < > > bzeeb-li...@lists.zabbadoz.net> wrote: > > > >> On 26 Nov 2020, at 4:29, Alan Somers wrote: > >> > >>> Author: asomers > >>> Date: Thu Nov 26 04:29:30 2020 > >>> New Revision: 368045 > >>> URL: https://svnweb.freebsd.org/changeset/base/368045 > >>> > >>> Log: > >>> Merge ping6 to ping > >>> > >>> There is now a single ping binary, which chooses to use ICMP or > >>> ICMPv4 > >>> based on the -4 and -6 options, and the format of the address. > >>> > >>> Submitted by: Ján Sučan > >>> Sponsored by: Google LLC (Google Summer of Code 2019) > >>> MFC after: Never > >>> Differential Revision: https://reviews.freebsd.org/D21377 > >> > >> I don’t have IPv4 anymore. > >> I don’t see any WITHOUT_INET or -DINET checks. > >> How can I compile INET out now? > >> > > > > I don't see any such checks before, either. Was it ever possible to > > exclude ping by building WITHOUT_INET? > > No, for various reasons including startup scripts relying on it etc. > I think no one ever bothered to fully find it all. > > But it was possible to just ditch the binary (not installing it into > custom images, or rm -f it post-install along with other things). > > > Now it’s a “dual-stack handling” binary and those we’ve tried > with a lot of care to make sure they grow compiling out both ways as you > cannot throw away the binary anymore. > > I don’t know how hard it is to do this now. I’ll be happy to have a > look and help if it’s not a 5 minute job for you knowing the code and > split up. Would we just have to handle main.c with #ifdefs and the old > ping[4] files from the Makefile or is ping6 also using shared code from > the former ping[4]? > > > /bz > Yes, that's about right. ping.c is the old ping.c, and ping6.c is the old ping6.c, with main() removed from both. So it should be pretty easy to compile out INET. -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368090 - stable/12/lib/libc/tests/iconv
Author: asomers
Date: Fri Nov 27 04:45:10 2020
New Revision: 368090
URL: https://svnweb.freebsd.org/changeset/base/368090
Log:
MFC r366132:
lib/libc/tests/iconv: raise WARNS to 6
Modified:
stable/12/lib/libc/tests/iconv/Makefile
stable/12/lib/libc/tests/iconv/iconvctl_test.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/lib/libc/tests/iconv/Makefile
==
--- stable/12/lib/libc/tests/iconv/Makefile Fri Nov 27 03:17:21 2020
(r368089)
+++ stable/12/lib/libc/tests/iconv/Makefile Fri Nov 27 04:45:10 2020
(r368090)
@@ -3,6 +3,5 @@
TESTSDIR= ${TESTSBASE}/lib/libc/iconv
ATF_TESTS_C+= iconvctl_test
-WARNS?=2
.include
Modified: stable/12/lib/libc/tests/iconv/iconvctl_test.c
==
--- stable/12/lib/libc/tests/iconv/iconvctl_test.c Fri Nov 27 03:17:21
2020(r368089)
+++ stable/12/lib/libc/tests/iconv/iconvctl_test.c Fri Nov 27 04:45:10
2020(r368090)
@@ -30,7 +30,7 @@ __FBSDID("$FreeBSD$");
#include
-void
+static void
test_trivialp(const char *src, const char *dst, int expected)
{
iconv_t ic;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368088 - stable/12/lib/libc/tests/sys
Author: asomers
Date: Thu Nov 26 23:54:13 2020
New Revision: 368088
URL: https://svnweb.freebsd.org/changeset/base/368088
Log:
MFC r366131:
lib/libc/tests/sys: raise WARNS to 6
Modified:
stable/12/lib/libc/tests/sys/Makefile
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/lib/libc/tests/sys/Makefile
==
--- stable/12/lib/libc/tests/sys/Makefile Thu Nov 26 23:44:57 2020
(r368087)
+++ stable/12/lib/libc/tests/sys/Makefile Thu Nov 26 23:54:13 2020
(r368088)
@@ -80,12 +80,6 @@ LIBADD.timer_create_test+= rt
SRCS.mlock_test+= mlock_helper.c
SRCS.setrlimit_test+= mlock_helper.c
-.if ${COMPILER_TYPE} == "gcc"
-WARNS?=3
-.else
-WARNS?=4
-.endif
-
FILESGROUPS+= truncate_test_FILES
truncate_test_FILES= truncate_test.root_owned
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368087 - stable/12/sys/kern
Author: asomers
Date: Thu Nov 26 23:44:57 2020
New Revision: 368087
URL: https://svnweb.freebsd.org/changeset/base/368087
Log:
MFC r366118:
Fix some signed/unsigned comparison warnings in NFS
Reviewed by: rmacklem
Differential Revision:https://reviews.freebsd.org/D26533
Modified:
stable/12/sys/kern/subr_acl_nfs4.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/kern/subr_acl_nfs4.c
==
--- stable/12/sys/kern/subr_acl_nfs4.c Thu Nov 26 23:35:13 2020
(r368086)
+++ stable/12/sys/kern/subr_acl_nfs4.c Thu Nov 26 23:44:57 2020
(r368087)
@@ -349,9 +349,9 @@ _acl_append(struct acl *aclp, acl_tag_t tag, acl_perm_
}
static struct acl_entry *
-_acl_duplicate_entry(struct acl *aclp, int entry_index)
+_acl_duplicate_entry(struct acl *aclp, unsigned entry_index)
{
- int i;
+ unsigned i;
KASSERT(aclp->acl_cnt + 1 <= ACL_MAX_ENTRIES,
("aclp->acl_cnt + 1 <= ACL_MAX_ENTRIES"));
@@ -368,7 +368,8 @@ static void
acl_nfs4_sync_acl_from_mode_draft(struct acl *aclp, mode_t mode,
int file_owner_id)
{
- int i, meets, must_append;
+ int meets, must_append;
+ unsigned i;
struct acl_entry *entry, *copy, *previous,
*a1, *a2, *a3, *a4, *a5, *a6;
mode_t amode;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368086 - stable/12/tools/regression/fsx
Author: asomers
Date: Thu Nov 26 23:35:13 2020
New Revision: 368086
URL: https://svnweb.freebsd.org/changeset/base/368086
Log:
MFC r365956:
fsx: fix build with WARNS=6
* signed/unsigned comparisons
* use standard warn(3)
* Suppress warnings about local vars and funcs not declared static
* const-correctness
* declaration shadows a variable in the global scope
Reviewed by: kevans
Differential Revision:https://reviews.freebsd.org/D26516
Modified:
stable/12/tools/regression/fsx/Makefile
stable/12/tools/regression/fsx/fsx.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/tools/regression/fsx/Makefile
==
--- stable/12/tools/regression/fsx/Makefile Thu Nov 26 23:34:02 2020
(r368085)
+++ stable/12/tools/regression/fsx/Makefile Thu Nov 26 23:35:13 2020
(r368086)
@@ -4,4 +4,9 @@ PROG= fsx
MAN=
+NO_WMISSING_VARIABLE_DECLARATIONS=
+
.include
+
+# Don't require static declarations. It's line noise in a single-file program.
+CWARNFLAGS+= -Wno-strict-prototypes -Wno-missing-prototypes
Modified: stable/12/tools/regression/fsx/fsx.c
==
--- stable/12/tools/regression/fsx/fsx.cThu Nov 26 23:34:02 2020
(r368085)
+++ stable/12/tools/regression/fsx/fsx.cThu Nov 26 23:35:13 2020
(r368086)
@@ -112,7 +112,7 @@ int closeprob = 0; /* -c flag */
intinvlprob = 0; /* -i flag */
intdebug = 0; /* -d flag */
unsigned long debugstart = 0; /* -D flag */
-unsigned long maxfilelen = 256 * 1024;/* -l flag */
+off_t maxfilelen = 256 * 1024;/* -l flag */
intsizechecks = 1; /* -n flag disables them */
intmaxoplen = 64 * 1024; /* -o flag */
intquiet = 0; /* -q flag */
@@ -138,33 +138,8 @@ int invl = 0;
void
-vwarnc(code, fmt, ap)
- int code;
- const char *fmt;
- va_list ap;
+prt(const char *fmt, ...)
{
- fprintf(stderr, "fsx: ");
- if (fmt != NULL) {
- vfprintf(stderr, fmt, ap);
- fprintf(stderr, ": ");
- }
- fprintf(stderr, "%s\n", strerror(code));
-}
-
-
-void
-warn(const char * fmt, ...)
-{
- va_list ap;
- va_start(ap, fmt);
- vwarnc(errno, fmt, ap);
- va_end(ap);
-}
-
-
-void
-prt(char *fmt, ...)
-{
va_list args;
va_start(args, fmt);
@@ -179,7 +154,7 @@ prt(char *fmt, ...)
}
void
-prterr(char *prefix)
+prterr(const char *prefix)
{
prt("%s%s%s\n", prefix, prefix ? ": " : "", strerror(errno));
}
@@ -317,12 +292,12 @@ logdump(void)
void
-save_buffer(char *buffer, off_t bufferlength, int fd)
+save_buffer(char *buffer, off_t bufferlength, int savefd)
{
off_t ret;
ssize_t byteswritten;
- if (fd <= 0 || bufferlength == 0)
+ if (savefd <= 0 || bufferlength == 0)
return;
if (bufferlength > SSIZE_MAX) {
@@ -330,7 +305,7 @@ save_buffer(char *buffer, off_t bufferlength, int fd)
exit(67);
}
if (lite) {
- off_t size_by_seek = lseek(fd, (off_t)0, SEEK_END);
+ off_t size_by_seek = lseek(savefd, (off_t)0, SEEK_END);
if (size_by_seek == (off_t)-1)
prterr("save_buffer: lseek eof");
else if (bufferlength > size_by_seek) {
@@ -340,11 +315,11 @@ save_buffer(char *buffer, off_t bufferlength, int fd)
}
}
- ret = lseek(fd, (off_t)0, SEEK_SET);
+ ret = lseek(savefd, (off_t)0, SEEK_SET);
if (ret == (off_t)-1)
prterr("save_buffer: lseek 0");
- byteswritten = write(fd, buffer, (size_t)bufferlength);
+ byteswritten = write(savefd, buffer, (size_t)bufferlength);
if (byteswritten != bufferlength) {
if (byteswritten == -1)
prterr("save_buffer write");
@@ -458,10 +433,10 @@ check_trunc_hack(void)
void
-doread(unsigned offset, unsigned size)
+doread(off_t offset, off_t size)
{
off_t ret;
- unsigned iret;
+ ssize_t iret;
offset -= offset % readbdy;
if (size == 0) {
@@ -509,7 +484,7 @@ doread(unsigned offset, unsigned size)
void
-check_eofpage(char *s, unsigned offset, char *p, int size)
+check_eofpage(const char *s, unsigned offset, char *p, int size)
{
uintptr_t last_page, should_be_zero;
@@ -592,7 +567,7 @@ domapread(unsigned offset, unsigned size)
void
-gendata(char *original_buf, char *good_buf, unsigned offset, unsigned size)
+gendata(unsigned offset, unsigned size)
{
while (size--) {
good_buf[offset] = testcalls % 256;
@@ -607,7 +582,7 @@ void
dowrite(unsigned offset, unsigned size)
{
off_t ret;
-
svn commit: r368085 - stable/12/lib/libc/gen
Author: asomers
Date: Thu Nov 26 23:34:02 2020
New Revision: 368085
URL: https://svnweb.freebsd.org/changeset/base/368085
Log:
MFC r365910:
fix integer underflow in getgrnam_r and getpwnam_r
Sometimes nscd(8) will return a 1-byte buffer for a nonexistent entry. This
triggered an integer underflow in grp_unmarshal_func, causing getgrnam_r to
return ERANGE instead of 0.
Fix the user's buffer size check, and add a correct check for a too-small
nscd buffer.
PR: 248932
Event:September 2020 Bugathon
Reviewed by: markj
Sponsored by: Axcient
Differential Revision: https://reviews.freebsd.org/D26204
Modified:
stable/12/lib/libc/gen/getgrent.c
stable/12/lib/libc/gen/getpwent.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/lib/libc/gen/getgrent.c
==
--- stable/12/lib/libc/gen/getgrent.c Thu Nov 26 22:49:05 2020
(r368084)
+++ stable/12/lib/libc/gen/getgrent.c Thu Nov 26 23:34:02 2020
(r368085)
@@ -334,14 +334,27 @@ grp_unmarshal_func(char *buffer, size_t buffer_size, v
orig_buf_size = va_arg(ap, size_t);
ret_errno = va_arg(ap, int *);
- if (orig_buf_size <
- buffer_size - sizeof(struct group) - sizeof(char *)) {
+ if (orig_buf_size + sizeof(struct group) + sizeof(char *) < buffer_size)
+ {
*ret_errno = ERANGE;
return (NS_RETURN);
+ } else if (buffer_size < sizeof(struct group) + sizeof(char *)) {
+ /*
+* nscd(8) sometimes returns buffer_size=1 for nonexistent
+* entries.
+*/
+ *ret_errno = 0;
+ return (NS_NOTFOUND);
}
memcpy(grp, buffer, sizeof(struct group));
memcpy(, buffer + sizeof(struct group), sizeof(char *));
+
+ if (orig_buf_size + sizeof(struct group) + sizeof(char *) +
+ _ALIGN(p) - (size_t)p < buffer_size) {
+ *ret_errno = ERANGE;
+ return (NS_RETURN);
+ }
orig_buf = (char *)_ALIGN(orig_buf);
memcpy(orig_buf, buffer + sizeof(struct group) + sizeof(char *) +
Modified: stable/12/lib/libc/gen/getpwent.c
==
--- stable/12/lib/libc/gen/getpwent.c Thu Nov 26 22:49:05 2020
(r368084)
+++ stable/12/lib/libc/gen/getpwent.c Thu Nov 26 23:34:02 2020
(r368085)
@@ -389,10 +389,17 @@ pwd_unmarshal_func(char *buffer, size_t buffer_size, v
orig_buf_size = va_arg(ap, size_t);
ret_errno = va_arg(ap, int *);
- if (orig_buf_size <
- buffer_size - sizeof(struct passwd) - sizeof(char *)) {
+ if (orig_buf_size + sizeof(struct passwd) + sizeof(char *) <
+ buffer_size) {
*ret_errno = ERANGE;
return (NS_RETURN);
+ } else if (buffer_size < sizeof(struct passwd) + sizeof(char *)) {
+ /*
+* nscd(8) sometimes returns buffer_size=1 for nonexistent
+* entries.
+*/
+ *ret_errno = 0;
+ return (NS_NOTFOUND);
}
memcpy(pwd, buffer, sizeof(struct passwd));
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r368045 - in head: . etc/mtree rescue/rescue sbin sbin/ping sbin/ping/tests sbin/ping6 sbin/ping6/tests tools/build/mk
On Thu, Nov 26, 2020 at 2:16 AM Bjoern A. Zeeb < bzeeb-li...@lists.zabbadoz.net> wrote: > On 26 Nov 2020, at 4:29, Alan Somers wrote: > > > Author: asomers > > Date: Thu Nov 26 04:29:30 2020 > > New Revision: 368045 > > URL: https://svnweb.freebsd.org/changeset/base/368045 > > > > Log: > > Merge ping6 to ping > > > > There is now a single ping binary, which chooses to use ICMP or > > ICMPv4 > > based on the -4 and -6 options, and the format of the address. > > > > Submitted by: Ján Sučan > > Sponsored by: Google LLC (Google Summer of Code 2019) > > MFC after: Never > > Differential Revision: https://reviews.freebsd.org/D21377 > > I don’t have IPv4 anymore. > I don’t see any WITHOUT_INET or -DINET checks. > How can I compile INET out now? > I don't see any such checks before, either. Was it ever possible to exclude ping by building WITHOUT_INET? > > > Also can we please have a ping6 [binary, compat shell script, or > hardlink] back which defaults to -6 with appropriate option parsing by > default? > People used that in scripts and whatnot for about 20 years (and > there’s even still stuff in tools in our own tree referencing it). > Done. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r368078 - in head: . rescue/rescue sbin/ping sbin/ping/tests tools/build/mk
Author: asomers
Date: Thu Nov 26 18:33:04 2020
New Revision: 368078
URL: https://svnweb.freebsd.org/changeset/base/368078
Log:
ping: add a ping6 hard link for backwards compatibility
When invoked as "ping6", ping will now attempt to use ICMPv6 for hostnames
that resolve both IPv4 and IPv6 addresses.
Reviewed by: bz, manu
MFC-With: r368045
Differential Revision:https://reviews.freebsd.org/D27384
Modified:
head/ObsoleteFiles.inc
head/rescue/rescue/Makefile
head/sbin/ping/Makefile
head/sbin/ping/main.c
head/sbin/ping/ping.8
head/sbin/ping/tests/ping_test.sh
head/tools/build/mk/OptionalObsoleteFiles.inc
Modified: head/ObsoleteFiles.inc
==
--- head/ObsoleteFiles.inc Thu Nov 26 18:16:32 2020(r368077)
+++ head/ObsoleteFiles.inc Thu Nov 26 18:33:04 2020(r368078)
@@ -37,8 +37,6 @@
# done
# 20201124: ping6(8) was merged into ping(8)
-OLD_FILES+=sbin/ping6
-OLD_FILES+=rescue/ping6
OLD_FILES+=usr/lib/debug/sbin/ping6.debug
OLD_FILES+=usr/share/man/man8/ping6.8.gz
OLD_FILES+=usr/tests/sbin/ping6/Kyuafile
Modified: head/rescue/rescue/Makefile
==
--- head/rescue/rescue/Makefile Thu Nov 26 18:16:32 2020(r368077)
+++ head/rescue/rescue/Makefile Thu Nov 26 18:33:04 2020(r368078)
@@ -103,6 +103,7 @@ CRUNCH_PROGS_sbin+= ccdconfig
.endif
.if ${MK_INET6_SUPPORT} != "no"
+CRUNCH_ALIAS_ping= ping6
CRUNCH_PROGS_sbin+= rtsol
.endif
Modified: head/sbin/ping/Makefile
==
--- head/sbin/ping/Makefile Thu Nov 26 18:16:32 2020(r368077)
+++ head/sbin/ping/Makefile Thu Nov 26 18:33:04 2020(r368078)
@@ -15,6 +15,7 @@ LIBADD= m
CFLAGS+= -DINET6 -DKAME_SCOPEID
SRCS+= ping6.c
LIBADD+= md
+LINKS= ${BINDIR}/ping ${BINDIR}/ping6
.endif
.if ${MK_DYNAMICROOT} == "no"
Modified: head/sbin/ping/main.c
==
--- head/sbin/ping/main.c Thu Nov 26 18:16:32 2020(r368077)
+++ head/sbin/ping/main.c Thu Nov 26 18:33:04 2020(r368078)
@@ -66,7 +66,10 @@ main(int argc, char *argv[])
struct in6_addr a6;
bool ipv6;
- ipv6 = false;
+ if (strcmp(getprogname(), "ping6") == 0)
+ ipv6 = true;
+ else
+ ipv6 = false;
#endif
ipv4 = false;
Modified: head/sbin/ping/ping.8
==
--- head/sbin/ping/ping.8 Thu Nov 26 18:16:32 2020(r368077)
+++ head/sbin/ping/ping.8 Thu Nov 26 18:33:04 2020(r368078)
@@ -28,7 +28,7 @@
.\" @(#)ping.8 8.2 (Berkeley) 12/11/93
.\" $FreeBSD$
.\"
-.Dd November 24, 2020
+.Dd November 26, 2020
.Dt PING 8
.Os
.Sh NAME
@@ -141,6 +141,9 @@ the specific IP version can be requested by
or
.Fl 6
options, respectively.
+For backwards-compatibility, ICMPv6 can also be selected by invoking the binary
+as
+.Nm ping6 .
.Ss Options common to both IPv4 and IPv6 targets
.Bl -tag -width indent
.It Fl A
Modified: head/sbin/ping/tests/ping_test.sh
==
--- head/sbin/ping/tests/ping_test.sh Thu Nov 26 18:16:32 2020
(r368077)
+++ head/sbin/ping/tests/ping_test.sh Thu Nov 26 18:33:04 2020
(r368078)
@@ -53,9 +53,23 @@ ping_6_c1_s8_t1_body() {
check_ping_statistics std.out $(atf_get_srcdir)/ping_6_c1_s8_t1.out
}
+atf_test_case ping6_c1_s8_t1
+ping6_c1_s8_t1_head() {
+atf_set "descr" "Use IPv6 when invoked as ping6"
+}
+ping6_c1_s8_t1_body() {
+if ! getaddrinfo -f inet6 localhost 1>/dev/null 2>&1; then
+ atf_skip "IPv6 is not configured"
+fi
+atf_check -s exit:0 -o save:std.out -e empty \
+ ping6 -c 1 -s 8 -t 1 localhost
+check_ping_statistics std.out $(atf_get_srcdir)/ping_6_c1_s8_t1.out
+}
+
atf_init_test_cases() {
atf_add_test_case ping_c1_s56_t1
atf_add_test_case ping_6_c1_s8_t1
+atf_add_test_case ping6_c1_s8_t1
}
check_ping_statistics() {
Modified: head/tools/build/mk/OptionalObsoleteFiles.inc
==
--- head/tools/build/mk/OptionalObsoleteFiles.inc Thu Nov 26 18:16:32
2020(r368077)
+++ head/tools/build/mk/OptionalObsoleteFiles.inc Thu Nov 26 18:33:04
2020(r368078)
@@ -2583,6 +2583,7 @@ OLD_DIRS+=usr/share/i18n/csmapper/GB
.endif
.if ${MK_INET6} == no
+OLD_FILES+=sbin/ping6
OLD_FILES+=sbin/rtsol
OLD_FILES+=usr/sbin/ip6addrctl
OLD_FILES+=usr/sbin/mld6query
@@ -2611,6 +2612,7 @@ OLD_FILES+=usr/share/man/man8/traceroute6.8.gz
.endif
.if ${MK_INET6_SUPPORT} == no
+OLD_FILES+=rescue/ping6
OLD_FILES+=rescue/rtsol
.endif
Re: svn commit: r368045 - in head: . etc/mtree rescue/rescue sbin sbin/ping sbin/ping/tests sbin/ping6 sbin/ping6/tests tools/build/mk
> > > > > I think you should add a LINKS=... here so people script which uses > > > ping6 won't break (and of course adding support in the code to do ipv6 > > > ping if progname is ping6 if this isn't the case). > > > > > > Cheers, > > > > > > -- > > > Emmanuel Vadot > > > > > > > Yes, that would make sense. But for how long? Would the ping6 hard link > > stick around forever, or eventually be removed in some future version of > > FreeBSD? > > I have no opinion on this matter. > But since ping6 was present for a very long time I guess we're stuck > with it for a long time too. > Do you have an opinion on whether there should be a /rescue/ping6 link too? ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r368045 - in head: . etc/mtree rescue/rescue sbin sbin/ping sbin/ping/tests sbin/ping6 sbin/ping6/tests tools/build/mk
On Thu, Nov 26, 2020 at 1:27 AM Emmanuel Vadot wrote: > > Hi Alan, > > On Thu, 26 Nov 2020 04:29:31 +0000 (UTC) > Alan Somers wrote: > > > Author: asomers > > Date: Thu Nov 26 04:29:30 2020 > > New Revision: 368045 > > URL: https://svnweb.freebsd.org/changeset/base/368045 > > > > Log: > > Merge ping6 to ping > > > > There is now a single ping binary, which chooses to use ICMP or ICMPv4 > > based on the -4 and -6 options, and the format of the address. > > > > Submitted by: Ján Su?an > > Sponsored by: Google LLC (Google Summer of Code 2019) > > MFC after: Never > > Differential Revision: https://reviews.freebsd.org/D21377 > > > > Added: > > head/sbin/ping/main.c (contents, props changed) > > head/sbin/ping/main.h (contents, props changed) > > head/sbin/ping/ping.h (contents, props changed) > > head/sbin/ping/ping6.c > > - copied, changed from r368010, head/sbin/ping6/ping6.c > > head/sbin/ping/ping6.h (contents, props changed) > > head/sbin/ping/tests/ping_6_c1_s8_t1.out > > - copied unchanged from r368010, > head/sbin/ping6/tests/ping6_c1_s8_t1.out > > Deleted: > > head/sbin/ping6/Makefile > > head/sbin/ping6/Makefile.depend > > head/sbin/ping6/ping6.8 > > head/sbin/ping6/ping6.c > > head/sbin/ping6/tests/Makefile > > head/sbin/ping6/tests/ping6_c1_s8_t1.out > > head/sbin/ping6/tests/ping6_test.sh > > Modified: > > head/ObsoleteFiles.inc > > head/UPDATING > > head/etc/mtree/BSD.tests.dist > > head/rescue/rescue/Makefile > > head/sbin/Makefile > > head/sbin/ping/Makefile > > head/sbin/ping/ping.8 > > head/sbin/ping/ping.c > > head/sbin/ping/tests/Makefile > > head/sbin/ping/tests/ping_test.sh > > head/tools/build/mk/OptionalObsoleteFiles.inc > > > > Modified: head/ObsoleteFiles.inc > > > == > > --- head/ObsoleteFiles.incThu Nov 26 02:14:52 2020(r368044) > > +++ head/ObsoleteFiles.incThu Nov 26 04:29:30 2020(r368045) > > @@ -36,6 +36,16 @@ > > # xargs -n1 | sort | uniq -d; > > # done > > > > +# 20201124: ping6(8) was merged into ping(8) > > +OLD_FILES+=sbin/ping6 > > +OLD_FILES+=rescue/ping6 > > +OLD_FILES+=usr/lib/debug/sbin/ping6.debug > > +OLD_FILES+=usr/share/man/man8/ping6.8.gz > > +OLD_FILES+=usr/tests/sbin/ping6/Kyuafile > > +OLD_FILES+=usr/tests/sbin/ping6/ping6_c1_s8_t1.out > > +OLD_FILES+=usr/tests/sbin/ping6/ping6_test > > +OLD_DIRS+=usr/tests/sbin/ping6 > > + > > # 20201025: Remove cal data files > > OLD_FILES+=usr/share/calendar/calendar.all > > OLD_FILES+=usr/share/calendar/calendar.australia > > > > Modified: head/UPDATING > > > == > > --- head/UPDATING Thu Nov 26 02:14:52 2020(r368044) > > +++ head/UPDATING Thu Nov 26 04:29:30 2020(r368045) > > @@ -26,6 +26,10 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW: > > world, or to merely disable the most expensive debugging > functionality > > at runtime, run "ln -s 'abort:false,junk:false' /etc/malloc.conf".) > > > > +20201124: > > + ping6 has been merged into ping. It can now be called as "ping > -6". > > + See ping(8) for details. > > + > > 20201108: > > Default value of net.add_addr_allfibs has been changed to 0. > > If you have multi-fib configuration and rely on existence of all > > > > Modified: head/etc/mtree/BSD.tests.dist > > > == > > --- head/etc/mtree/BSD.tests.dist Thu Nov 26 02:14:52 2020 > (r368044) > > +++ head/etc/mtree/BSD.tests.dist Thu Nov 26 04:29:30 2020 > (r368045) > > @@ -448,8 +448,6 @@ > > .. > > ping > > .. > > -ping6 > > -.. > > route > > .. > > .. > > > > Modified: head/rescue/rescue/Makefile > > > == > > --- head/rescue/rescue/Makefile Thu Nov 26 02:14:52 2020 > (r368044) > > +++ head/rescue/rescue/Makefile Thu Nov 26 04:29:30 2020 > (r368045) > > @@ -103,7 +103,6 @@ CRUNCH_PROGS_sbin+= ccdconfig > > .endif > > > >
svn commit: r368046 - in head: contrib/traceroute share/man/man4 tests/sys/netinet tests/sys/netinet6 tests/sys/netipsec/tunnel tests/sys/netpfil/common tests/sys/netpfil/pf usr.sbin/traceroute6
Author: asomers
Date: Thu Nov 26 04:55:02 2020
New Revision: 368046
URL: https://svnweb.freebsd.org/changeset/base/368046
Log:
ping: fix some man pages and tests after r368045
MFC-with: r368045
Modified:
head/contrib/traceroute/traceroute.8
head/share/man/man4/inet6.4
head/tests/sys/netinet/carp.sh
head/tests/sys/netinet6/exthdr.sh
head/tests/sys/netinet6/lpm6.sh
head/tests/sys/netinet6/output6.sh
head/tests/sys/netipsec/tunnel/utils.subr
head/tests/sys/netpfil/common/pass_block.sh
head/tests/sys/netpfil/pf/fragmentation.sh
head/tests/sys/netpfil/pf/pass_block.sh
head/tests/sys/netpfil/pf/table.sh
head/usr.sbin/traceroute6/traceroute6.8
Modified: head/contrib/traceroute/traceroute.8
==
--- head/contrib/traceroute/traceroute.8Thu Nov 26 04:29:30 2020
(r368045)
+++ head/contrib/traceroute/traceroute.8Thu Nov 26 04:55:02 2020
(r368046)
@@ -16,7 +16,7 @@
.\"$Id: traceroute.8,v 1.19 2000/09/21 08:44:19 leres Exp $
.\"$FreeBSD$
.\"
-.Dd June 20, 2019
+.Dd November 25, 2020
.Dt TRACEROUTE 8
.Os
.Sh NAME
@@ -377,7 +377,6 @@ during normal operations or from automated scripts.
.Sh SEE ALSO
.Xr netstat 1 ,
.Xr ping 8 ,
-.Xr ping6 8 ,
.Xr traceroute6 8 .
.Sh AUTHORS
Implemented by Van Jacobson from a suggestion by Steve Deering. Debugged
Modified: head/share/man/man4/inet6.4
==
--- head/share/man/man4/inet6.4 Thu Nov 26 04:29:30 2020(r368045)
+++ head/share/man/man4/inet6.4 Thu Nov 26 04:55:02 2020(r368046)
@@ -29,7 +29,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd August 14, 2018
+.Dd November 25, 2020
.Dt INET6 4
.Os
.Sh NAME
@@ -130,7 +130,7 @@ or
are able to use this notation.
With special programs
like
-.Xr ping6 8 ,
+.Xr ping 8 ,
you can specify the outgoing interface by an extra command line option
to disambiguate scoped addresses.
.Pp
Modified: head/tests/sys/netinet/carp.sh
==
--- head/tests/sys/netinet/carp.sh Thu Nov 26 04:29:30 2020
(r368045)
+++ head/tests/sys/netinet/carp.sh Thu Nov 26 04:55:02 2020
(r368046)
@@ -137,7 +137,7 @@ basic_v6_body()
carp_basic_v6_three ${epair_two}b
atf_check -s exit:0 -o ignore jexec carp_basic_v6_one \
- ping6 -c 3 2001:db8::0:1
+ ping -6 -c 3 2001:db8::0:1
}
basic_v6_cleanup()
Modified: head/tests/sys/netinet6/exthdr.sh
==
--- head/tests/sys/netinet6/exthdr.sh Thu Nov 26 04:29:30 2020
(r368045)
+++ head/tests/sys/netinet6/exthdr.sh Thu Nov 26 04:55:02 2020
(r368046)
@@ -74,7 +74,7 @@ exthdr_body() {
pyname=$(atf_get ident)
pyname=${pyname%*_[0-9]}
- atf_check -o ignore -s exit:0 ping6 -c 3 -q -o ${ip6b}
+ atf_check -o ignore -s exit:0 ping -6 -c 3 -q -o ${ip6b}
atf_check -s exit:0 $(atf_get_srcdir)/${pyname}.py \
--sendif ${epair}a --recvif ${epair}a \
Modified: head/tests/sys/netinet6/lpm6.sh
==
--- head/tests/sys/netinet6/lpm6.sh Thu Nov 26 04:29:30 2020
(r368045)
+++ head/tests/sys/netinet6/lpm6.sh Thu Nov 26 04:55:02 2020
(r368046)
@@ -100,7 +100,7 @@ lpm6_test1_success_body()
valid_message="${count} packets transmitted, ${count} packets received"
# Check that ${net_dst}:2:0 goes via epair0
- atf_check -o match:"${valid_message}" jexec ${jname}a ping6 -f
-nc${count} ${net_dst}:2:0
+ atf_check -o match:"${valid_message}" jexec ${jname}a ping -6 -f
-nc${count} ${net_dst}:2:0
pkt_0=`jexec ${jname}a netstat -Wf link -I ${epair0}a | head | awk
'$1!~/^Name/{print$8}'`
pkt_1=`jexec ${jname}a netstat -Wf link -I ${epair1}a | head | awk
'$1!~/^Name/{print$8}'`
if [ ${pkt_0} -le ${count} ]; then
@@ -109,7 +109,7 @@ lpm6_test1_success_body()
fi
# Check that ${net_dst}:2:1 goes via epair1
- atf_check -o match:"${valid_message}" jexec ${jname}a ping6 -f
-nc${count} ${net_dst}:2:1
+ atf_check -o match:"${valid_message}" jexec ${jname}a ping -6 -f
-nc${count} ${net_dst}:2:1
pkt_0=`jexec ${jname}a netstat -Wf link -I ${epair0}a | head | awk
'$1!~/^Name/{print$8}'`
pkt_1=`jexec ${jname}a netstat -Wf link -I ${epair1}a | head | awk
'$1!~/^Name/{print$8}'`
if [ ${pkt_1} -le ${count} ]; then
@@ -163,7 +163,7 @@ lpm6_test2_success_body()
valid_message="${count} packets transmitted, ${count} packets received"
# Check that ${net_dst}:2:0 goes via epair1
- atf_check -o match:"${valid_message}" jexec ${jname}a ping6 -f
-nc${count} ${net_dst}:2:0
+ atf_check -o match:"${valid_message}"
svn commit: r368045 - in head: . etc/mtree rescue/rescue sbin sbin/ping sbin/ping/tests sbin/ping6 sbin/ping6/tests tools/build/mk
Author: asomers
Date: Thu Nov 26 04:29:30 2020
New Revision: 368045
URL: https://svnweb.freebsd.org/changeset/base/368045
Log:
Merge ping6 to ping
There is now a single ping binary, which chooses to use ICMP or ICMPv4
based on the -4 and -6 options, and the format of the address.
Submitted by: Ján Sučan
Sponsored by: Google LLC (Google Summer of Code 2019)
MFC after:Never
Differential Revision:https://reviews.freebsd.org/D21377
Added:
head/sbin/ping/main.c (contents, props changed)
head/sbin/ping/main.h (contents, props changed)
head/sbin/ping/ping.h (contents, props changed)
head/sbin/ping/ping6.c
- copied, changed from r368010, head/sbin/ping6/ping6.c
head/sbin/ping/ping6.h (contents, props changed)
head/sbin/ping/tests/ping_6_c1_s8_t1.out
- copied unchanged from r368010, head/sbin/ping6/tests/ping6_c1_s8_t1.out
Deleted:
head/sbin/ping6/Makefile
head/sbin/ping6/Makefile.depend
head/sbin/ping6/ping6.8
head/sbin/ping6/ping6.c
head/sbin/ping6/tests/Makefile
head/sbin/ping6/tests/ping6_c1_s8_t1.out
head/sbin/ping6/tests/ping6_test.sh
Modified:
head/ObsoleteFiles.inc
head/UPDATING
head/etc/mtree/BSD.tests.dist
head/rescue/rescue/Makefile
head/sbin/Makefile
head/sbin/ping/Makefile
head/sbin/ping/ping.8
head/sbin/ping/ping.c
head/sbin/ping/tests/Makefile
head/sbin/ping/tests/ping_test.sh
head/tools/build/mk/OptionalObsoleteFiles.inc
Modified: head/ObsoleteFiles.inc
==
--- head/ObsoleteFiles.inc Thu Nov 26 02:14:52 2020(r368044)
+++ head/ObsoleteFiles.inc Thu Nov 26 04:29:30 2020(r368045)
@@ -36,6 +36,16 @@
# xargs -n1 | sort | uniq -d;
# done
+# 20201124: ping6(8) was merged into ping(8)
+OLD_FILES+=sbin/ping6
+OLD_FILES+=rescue/ping6
+OLD_FILES+=usr/lib/debug/sbin/ping6.debug
+OLD_FILES+=usr/share/man/man8/ping6.8.gz
+OLD_FILES+=usr/tests/sbin/ping6/Kyuafile
+OLD_FILES+=usr/tests/sbin/ping6/ping6_c1_s8_t1.out
+OLD_FILES+=usr/tests/sbin/ping6/ping6_test
+OLD_DIRS+=usr/tests/sbin/ping6
+
# 20201025: Remove cal data files
OLD_FILES+=usr/share/calendar/calendar.all
OLD_FILES+=usr/share/calendar/calendar.australia
Modified: head/UPDATING
==
--- head/UPDATING Thu Nov 26 02:14:52 2020(r368044)
+++ head/UPDATING Thu Nov 26 04:29:30 2020(r368045)
@@ -26,6 +26,10 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW:
world, or to merely disable the most expensive debugging functionality
at runtime, run "ln -s 'abort:false,junk:false' /etc/malloc.conf".)
+20201124:
+ ping6 has been merged into ping. It can now be called as "ping -6".
+ See ping(8) for details.
+
20201108:
Default value of net.add_addr_allfibs has been changed to 0.
If you have multi-fib configuration and rely on existence of all
Modified: head/etc/mtree/BSD.tests.dist
==
--- head/etc/mtree/BSD.tests.dist Thu Nov 26 02:14:52 2020
(r368044)
+++ head/etc/mtree/BSD.tests.dist Thu Nov 26 04:29:30 2020
(r368045)
@@ -448,8 +448,6 @@
..
ping
..
-ping6
-..
route
..
..
Modified: head/rescue/rescue/Makefile
==
--- head/rescue/rescue/Makefile Thu Nov 26 02:14:52 2020(r368044)
+++ head/rescue/rescue/Makefile Thu Nov 26 04:29:30 2020(r368045)
@@ -103,7 +103,6 @@ CRUNCH_PROGS_sbin+= ccdconfig
.endif
.if ${MK_INET6_SUPPORT} != "no"
-CRUNCH_PROGS_sbin+= ping6
CRUNCH_PROGS_sbin+= rtsol
.endif
Modified: head/sbin/Makefile
==
--- head/sbin/Makefile Thu Nov 26 02:14:52 2020(r368044)
+++ head/sbin/Makefile Thu Nov 26 04:29:30 2020(r368045)
@@ -73,7 +73,6 @@ SUBDIR.${MK_CCD}+=ccdconfig
SUBDIR.${MK_CXX}+= devd
SUBDIR.${MK_HAST}+=hastctl
SUBDIR.${MK_HAST}+=hastd
-SUBDIR.${MK_INET6}+= ping6
SUBDIR.${MK_INET6}+= rtsol
SUBDIR.${MK_IPFILTER}+=ipf
SUBDIR.${MK_IPFW}+=ipfw
Modified: head/sbin/ping/Makefile
==
--- head/sbin/ping/Makefile Thu Nov 26 02:14:52 2020(r368044)
+++ head/sbin/ping/Makefile Thu Nov 26 04:29:30 2020(r368045)
@@ -5,11 +5,17 @@
PACKAGE=runtime
PROG= ping
-SRCS= ping.c utils.c
+SRCS= main.c ping.c utils.c
MAN= ping.8
BINOWN=root
BINMODE=4555
LIBADD=m
+
+.if ${MK_INET6_SUPPORT} != "no"
+CFLAGS+= -DINET6 -DKAME_SCOPEID
+SRCS+= ping6.c
+LIBADD+= md
+.endif
.if ${MK_DYNAMICROOT} == "no"
.warning ${PROG} built without libcasper support
Added: head/sbin/ping/main.c
svn commit: r367976 - head/sbin/ping6
Author: asomers Date: Tue Nov 24 02:51:45 2020 New Revision: 367976 URL: https://svnweb.freebsd.org/changeset/base/367976 Log: ping6: update usage text after r365547 MFC after:2 weeks Modified: head/sbin/ping6/ping6.c Modified: head/sbin/ping6/ping6.c == --- head/sbin/ping6/ping6.c Tue Nov 24 02:05:43 2020(r367975) +++ head/sbin/ping6/ping6.c Tue Nov 24 02:51:45 2020(r367976) @@ -2868,7 +2868,8 @@ usage(void) " [-P policy]" #endif " [-S sourceaddr] [-s packetsize]\n" - " [-t timeout] [-W waittime] [hops ...] host\n"); + " [-t timeout] [-W waittime] [-z tclass] [hops ...] " + "host\n"); exit(1); } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r367785 - in head/sys: fs/nfs nfsserver
Author: asomers
Date: Wed Nov 18 04:35:49 2020
New Revision: 367785
URL: https://svnweb.freebsd.org/changeset/base/367785
Log:
nfs: Mark unused statistics variable as reserved
FreeBSD's NFS exporter has long exported some unused statistics fields.
Revision r366992 removed them from nfsstat. This revision renames those
fields in the kernel's exported structures to make it clear to other
consumers that they are unused.
Reported by: emaste
Reviewed by: emaste
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D27258
Modified:
head/sys/fs/nfs/nfs_commonport.c
head/sys/fs/nfs/nfsport.h
head/sys/nfsserver/nfsrvstats.h
Modified: head/sys/fs/nfs/nfs_commonport.c
==
--- head/sys/fs/nfs/nfs_commonport.cWed Nov 18 03:43:03 2020
(r367784)
+++ head/sys/fs/nfs/nfs_commonport.cWed Nov 18 04:35:49 2020
(r367785)
@@ -541,16 +541,15 @@ nfssvc_call(struct thread *p, struct nfssvc_args *uap,
i < NFSV42_NOPS + NFSV4OP_FAKENOPS; i++, j++)
oldnfsstats.srvrpccnt[j] =
nfsstatsv1.srvrpccnt[i];
- oldnfsstats.srvrpc_errs = nfsstatsv1.srvrpc_errs;
- oldnfsstats.srv_errs = nfsstatsv1.srv_errs;
+ oldnfsstats.reserved_0 = 0;
+ oldnfsstats.reserved_1 = 0;
oldnfsstats.rpcrequests = nfsstatsv1.rpcrequests;
oldnfsstats.rpctimeouts = nfsstatsv1.rpctimeouts;
oldnfsstats.rpcunexpected = nfsstatsv1.rpcunexpected;
oldnfsstats.rpcinvalid = nfsstatsv1.rpcinvalid;
oldnfsstats.srvcache_inproghits =
nfsstatsv1.srvcache_inproghits;
- oldnfsstats.srvcache_idemdonehits =
- nfsstatsv1.srvcache_idemdonehits;
+ oldnfsstats.reserved_2 = 0;
oldnfsstats.srvcache_nonidemdonehits =
nfsstatsv1.srvcache_nonidemdonehits;
oldnfsstats.srvcache_misses =
@@ -636,10 +635,8 @@ nfssvc_call(struct thread *p, struct nfssvc_args *uap,
i++, j++)
nfsstatsov1.srvrpccnt[j] =
nfsstatsv1.srvrpccnt[i];
- nfsstatsov1.srvrpc_errs =
- nfsstatsv1.srvrpc_errs;
- nfsstatsov1.srv_errs =
- nfsstatsv1.srv_errs;
+ nfsstatsov1.reserved_0 = 0;
+ nfsstatsov1.reserved_1 = 0;
nfsstatsov1.rpcrequests =
nfsstatsv1.rpcrequests;
nfsstatsov1.rpctimeouts =
@@ -650,8 +647,7 @@ nfssvc_call(struct thread *p, struct nfssvc_args *uap,
nfsstatsv1.rpcinvalid;
nfsstatsov1.srvcache_inproghits =
nfsstatsv1.srvcache_inproghits;
- nfsstatsov1.srvcache_idemdonehits =
- nfsstatsv1.srvcache_idemdonehits;
+ nfsstatsov1.reserved_2 = 0;
nfsstatsov1.srvcache_nonidemdonehits =
nfsstatsv1.srvcache_nonidemdonehits;
nfsstatsov1.srvcache_misses =
@@ -750,10 +746,7 @@ nfssvc_call(struct thread *p, struct nfssvc_args *uap,
sizeof(nfsstatsv1.rpccnt));
}
if ((uap->flag & NFSSVC_ZEROSRVSTATS) != 0) {
- nfsstatsv1.srvrpc_errs = 0;
- nfsstatsv1.srv_errs = 0;
nfsstatsv1.srvcache_inproghits = 0;
- nfsstatsv1.srvcache_idemdonehits = 0;
nfsstatsv1.srvcache_nonidemdonehits = 0;
nfsstatsv1.srvcache_misses = 0;
nfsstatsv1.srvcache_tcppeak = 0;
Modified: head/sys/fs/nfs/nfsport.h
==
--- head/sys/fs/nfs/nfsport.h Wed Nov 18 03:43:03 2020(r367784)
+++ head/sys/fs/nfs/nfsport.h Wed Nov 18 04:35:49 2020(r367785)
@@ -447,14 +447,14 @@ struct nfsstatsv1 {
uint64_trpccnt[NFSV42_NPROCS + 15];
uint64_trpcretries;
uint64_t
svn commit: r366992 - head/usr.bin/nfsstat
Author: asomers
Date: Sat Oct 24 05:52:29 2020
New Revision: 366992
URL: https://svnweb.freebsd.org/changeset/base/366992
Log:
nfsstat: delete unused fields
Ever since r192762 nfsstat has included a few fields whose values were
always 0. They were copied from OpenBSD, but have never been used on
FreeBSD. Don't display them.
Reviewed by: rmacklem
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D26920
Modified:
head/usr.bin/nfsstat/nfsstat.c
Modified: head/usr.bin/nfsstat/nfsstat.c
==
--- head/usr.bin/nfsstat/nfsstat.c Sat Oct 24 05:26:58 2020
(r366991)
+++ head/usr.bin/nfsstat/nfsstat.c Sat Oct 24 05:52:29 2020
(r366992)
@@ -450,12 +450,7 @@ intpr(int clientOnly, int serverOnly)
xo_close_container("operations");
xo_open_container("server");
- xo_emit("{T:Server Re-Failed}\n");
- xo_emit("{:retfailed/%16ju}\n",
(uintmax_t)ext_nfsstats.srvrpc_errs);
- xo_emit("{T:Server Faults}\n");
- xo_emit("{:faults/%13ju}\n", (uintmax_t)ext_nfsstats.srv_errs);
-
xo_emit("{T:Server Write Gathering:/%13.13s}\n");
xo_emit("{T:WriteOps/%13.13s}{T:WriteRPC/%13.13s}"
@@ -473,12 +468,10 @@ intpr(int clientOnly, int serverOnly)
xo_open_container("cache");
xo_emit("{T:Server Cache Stats:/%13.13s}\n");
- xo_emit("{T:Inprog/%13.13s}{T:Idem/%13.13s}"
+ xo_emit("{T:Inprog/%13.13s}"
"{T:Non-Idem/%13.13s}{T:Misses/%13.13s}\n");
- xo_emit("{:inprog/%13ju}{:idem/%13ju}"
- "{:nonidem/%13ju}{:misses/%13ju}\n",
+ xo_emit("{:inprog/%13ju}{:nonidem/%13ju}{:misses/%13ju}\n",
(uintmax_t)ext_nfsstats.srvcache_inproghits,
- (uintmax_t)ext_nfsstats.srvcache_idemdonehits,
(uintmax_t)ext_nfsstats.srvcache_nonidemdonehits,
(uintmax_t)ext_nfsstats.srvcache_misses);
xo_close_container("cache");
@@ -1057,17 +1050,12 @@ exp_intpr(int clientOnly, int serverOnly, int nfs41)
if (printtitle)
xo_emit("{T:Server:}\n");
xo_open_container("server");
- xo_emit("{T:Retfailed/%13.13s}{T:Faults/%13.13s}"
- "{T:Clients/%13.13s}\n");
- xo_emit("{:retfailed/%13ju}{:faults/%13ju}{:clients/%13ju}\n",
- (uintmax_t)ext_nfsstats.srv_errs,
- (uintmax_t)ext_nfsstats.srvrpc_errs,
- (uintmax_t)ext_nfsstats.srvclients);
- xo_emit("{T:OpenOwner/%13.13s}{T:Opens/%13.13s}"
- "{T:LockOwner/%13.13s}{T:Locks/%13.13s}"
+ xo_emit("{T:Clients/%13.13s}{T:OpenOwner/%13.13s}"
+ "{T:Opens/%13.13s}{T:LockOwner/%13.13s}{T:Locks/%13.13s}"
"{T:Delegs/%13.13s}\n");
- xo_emit("{:openowner/%13ju}{:opens/%13ju}{:lockowner/%13ju}"
- "{:locks/%13ju}{:delegs/%13ju}\n",
+ xo_emit("{:clients/%13ju}{:openowner/%13ju}{:opens/%13ju}"
+ "{:lockowner/%13ju}{:locks/%13ju}{:delegs/%13ju}\n",
+ (uintmax_t)ext_nfsstats.srvclients,
(uintmax_t)ext_nfsstats.srvopenowners,
(uintmax_t)ext_nfsstats.srvopens,
(uintmax_t)ext_nfsstats.srvlockowners,
@@ -1078,13 +1066,12 @@ exp_intpr(int clientOnly, int serverOnly, int nfs41)
if (printtitle)
xo_emit("{T:Server Cache Stats:}\n");
xo_open_container("cache");
- xo_emit("{T:Inprog/%13.13s}{T:Idem/%13.13s}"
+ xo_emit("{T:Inprog/%13.13s}"
"{T:Non-idem/%13.13s}{T:Misses/%13.13s}"
"{T:CacheSize/%13.13s}{T:TCPPeak/%13.13s}\n");
- xo_emit("{:inprog/%13ju}{:idem/%13ju}{:nonidem/%13ju}"
+ xo_emit("{:inprog/%13ju}{:nonidem/%13ju}"
"{:misses/%13ju}{:cachesize/%13ju}{:tcppeak/%13ju}\n",
(uintmax_t)ext_nfsstats.srvcache_inproghits,
- (uintmax_t)ext_nfsstats.srvcache_idemdonehits,
(uintmax_t)ext_nfsstats.srvcache_nonidemdonehits,
(uintmax_t)ext_nfsstats.srvcache_misses,
(uintmax_t)ext_nfsstats.srvcache_size,
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r366365 - head/tests/sys/fs/fusefs
Author: asomers
Date: Fri Oct 2 17:06:05 2020
New Revision: 366365
URL: https://svnweb.freebsd.org/changeset/base/366365
Log:
fusefs tests: quell Coverity "Argument cannot be negative" warnings
Must abort tests early if open(2) fails.
Reported by: Coverity
Coverity CID: 1432810 and many others
Reviewed by: kevans
MFC after:2 weeks
Differential Revision:https://reviews.freebsd.org/D26635
Modified:
head/tests/sys/fs/fusefs/allow_other.cc
head/tests/sys/fs/fusefs/create.cc
head/tests/sys/fs/fusefs/default_permissions.cc
head/tests/sys/fs/fusefs/flush.cc
head/tests/sys/fs/fusefs/open.cc
head/tests/sys/fs/fusefs/opendir.cc
head/tests/sys/fs/fusefs/release.cc
head/tests/sys/fs/fusefs/releasedir.cc
head/tests/sys/fs/fusefs/write.cc
Modified: head/tests/sys/fs/fusefs/allow_other.cc
==
--- head/tests/sys/fs/fusefs/allow_other.cc Fri Oct 2 15:37:51 2020
(r366364)
+++ head/tests/sys/fs/fusefs/allow_other.cc Fri Oct 2 17:06:05 2020
(r366365)
@@ -168,7 +168,7 @@ TEST_F(AllowOther, privilege_escalation)
.WillRepeatedly(Invoke(ReturnErrno(EPERM)));
fd1 = open(FULLPATH, O_RDONLY);
- EXPECT_LE(0, fd1) << strerror(errno);
+ ASSERT_LE(0, fd1) << strerror(errno);
}, [] {
int fd0;
Modified: head/tests/sys/fs/fusefs/create.cc
==
--- head/tests/sys/fs/fusefs/create.cc Fri Oct 2 15:37:51 2020
(r366364)
+++ head/tests/sys/fs/fusefs/create.cc Fri Oct 2 17:06:05 2020
(r366365)
@@ -143,7 +143,7 @@ TEST_F(Create, attr_cache)
).Times(0);
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -184,7 +184,7 @@ TEST_F(Create, clear_attr_cache)
EXPECT_EQ(0, stat("mountpoint", )) << strerror(errno);
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
EXPECT_EQ(0, stat("mountpoint", )) << strerror(errno);
leak(fd);
@@ -254,7 +254,7 @@ TEST_F(Create, Enosys)
})));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -366,7 +366,7 @@ TEST_F(Create, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -398,7 +398,7 @@ TEST_F(Create, wronly_0444)
}));
fd = open(FULLPATH, O_CREAT | O_WRONLY, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -422,7 +422,7 @@ TEST_F(Create_7_8, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -446,6 +446,6 @@ TEST_F(Create_7_11, ok)
}));
fd = open(FULLPATH, O_CREAT | O_EXCL, mode);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
Modified: head/tests/sys/fs/fusefs/default_permissions.cc
==
--- head/tests/sys/fs/fusefs/default_permissions.cc Fri Oct 2 15:37:51
2020(r366364)
+++ head/tests/sys/fs/fusefs/default_permissions.cc Fri Oct 2 17:06:05
2020(r366365)
@@ -490,7 +490,7 @@ TEST_F(Create, ok)
expect_create(RELPATH, ino);
fd = open(FULLPATH, O_CREAT | O_EXCL, 0644);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
@@ -765,7 +765,7 @@ TEST_F(Open, ok)
expect_open(ino, 0, 1);
fd = open(FULLPATH, O_RDONLY);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
leak(fd);
}
Modified: head/tests/sys/fs/fusefs/flush.cc
==
--- head/tests/sys/fs/fusefs/flush.cc Fri Oct 2 15:37:51 2020
(r366364)
+++ head/tests/sys/fs/fusefs/flush.cc Fri Oct 2 17:06:05 2020
(r366365)
@@ -102,10 +102,10 @@ TEST_F(Flush, open_twice)
expect_release();
fd = open(FULLPATH, O_WRONLY);
- EXPECT_LE(0, fd) << strerror(errno);
+ ASSERT_LE(0, fd) << strerror(errno);
fd2 = open(FULLPATH, O_WRONLY);
- EXPECT_LE(0, fd2) << strerror(errno);
+ ASSERT_LE(0, fd2) << strerror(errno);
EXPECT_EQ(0, close(fd2)) << strerror(errno);
EXPECT_EQ(0, close(fd)) << strerror(errno);
@@ -132,7 +132,7 @@ TEST_F(Flush, eio)
svn commit: r366341 - stable/12/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Thu Oct 1 19:55:52 2020
New Revision: 366341
URL: https://svnweb.freebsd.org/changeset/base/366341
Log:
zfs: fix "zfs receive" of interrupted stream without "-F" after r366180
The OpenZFS test suite revealed a regression with the changes made by
r366180 and r364974: "zfs recv" resuming and interrupted stream without -F
would refuse to run, complaining that "dataset exists".
Direct commit to stable/12 because head has moved to the OpenZFS code base.
Upstream issue: https://github.com/openzfs/zfs/issues/10995
Reviewed by: mmacy
MFC after:3 days
MFC-with: 366180
Sponsored by: Axcient
Modified:
stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Modified: stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Thu Oct 1 19:17:03 2020(r366340)
+++ stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Thu Oct 1 19:55:52 2020(r366341)
@@ -3138,7 +3138,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
char prop_errbuf[1024];
const char *chopprefix;
boolean_t newfs = B_FALSE;
- boolean_t stream_wantsnewfs;
+ boolean_t stream_wantsnewfs, stream_resumingnewfs;
uint64_t parent_snapguid = 0;
prop_changelist_t *clp = NULL;
nvlist_t *snapprops_nvlist = NULL;
@@ -3301,7 +3301,9 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
boolean_t resuming = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_RESUMING;
stream_wantsnewfs = (drrb->drr_fromguid == 0 ||
- (drrb->drr_flags & DRR_FLAG_CLONE) || originsnap);
+ (drrb->drr_flags & DRR_FLAG_CLONE) || originsnap) && !resuming;
+ stream_resumingnewfs = (drrb->drr_fromguid == 0 ||
+ (drrb->drr_flags & DRR_FLAG_CLONE) || originsnap) && resuming;
if (stream_wantsnewfs) {
/*
@@ -3433,7 +3435,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
- stream_wantsnewfs) {
+ (stream_wantsnewfs || stream_resumingnewfs)) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, 0);
if (clp == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r366207 - head/lib/libc/gen
On Sun, Sep 27, 2020 at 5:15 PM Konstantin Belousov
wrote:
> On Sun, Sep 27, 2020 at 10:26:41PM +0000, Alan Somers wrote:
> > Author: asomers
> > Date: Sun Sep 27 22:26:41 2020
> > New Revision: 366207
> > URL: https://svnweb.freebsd.org/changeset/base/366207
> >
> > Log:
> > Misc compiler warning fixes in lib/libc
> >
> > Reviewed by:kevans, imp
> > MFC after: 2 weeks
> > Differential Revision: https://reviews.freebsd.org/D26534
> >
> > Modified:
> > head/lib/libc/gen/auxv.c
> > head/lib/libc/gen/basename_compat.c
> > head/lib/libc/gen/crypt.c
> > head/lib/libc/gen/dirname_compat.c
> > head/lib/libc/gen/fts-compat.c
> > head/lib/libc/gen/ftw-compat11.c
> > head/lib/libc/gen/getentropy.c
> >
> > Modified: head/lib/libc/gen/auxv.c
> >
> ==
> > --- head/lib/libc/gen/auxv.c Sun Sep 27 21:43:19 2020(r366206)
> > +++ head/lib/libc/gen/auxv.c Sun Sep 27 22:26:41 2020(r366207)
> > @@ -67,7 +67,8 @@ __init_elf_aux_vector(void)
> > }
> >
> > static pthread_once_t aux_once = PTHREAD_ONCE_INIT;
> > -static int pagesize, osreldate, canary_len, ncpus, pagesizes_len,
> bsdflags;
> > +static int pagesize, osreldate, ncpus, bsdflags;
> > +static size_t canary_len, pagesizes_len;
> > static int hwcap_present, hwcap2_present;
> > static char *canary, *pagesizes, *execpath;
> > static void *ps_strings, *timekeep;
> > @@ -245,16 +246,21 @@ int
> > _elf_aux_info(int aux, void *buf, int buflen)
> > {
> > int res;
> > + size_t buflen_;
> >
> > __init_elf_aux_vector();
> > if (__elf_aux_vector == NULL)
> > return (ENOSYS);
> > _once(_once, init_aux);
> >
> > + if (buflen < 0)
> > + return (EINVAL);
> > + buflen_ = (size_t)buflen;
> > +
> > switch (aux) {
> > case AT_CANARY:
> > - if (canary != NULL && canary_len >= buflen) {
> > - memcpy(buf, canary, buflen);
> > + if (canary != NULL && canary_len >= buflen_) {
> > + memcpy(buf, canary, buflen_);
> > memset(canary, 0, canary_len);
> > canary = NULL;
> > res = 0;
> > @@ -267,35 +273,35 @@ _elf_aux_info(int aux, void *buf, int buflen)
> > else if (buf == NULL)
> > res = EINVAL;
> > else {
> > - if (strlcpy(buf, execpath, buflen) >= buflen)
> > + if (strlcpy(buf, execpath, buflen_) >= buflen_)
> > res = EINVAL;
> > else
> > res = 0;
> > }
> > break;
> > case AT_HWCAP:
> > - if (hwcap_present && buflen == sizeof(u_long)) {
> > + if (hwcap_present && buflen_ == sizeof(u_long)) {
> > *(u_long *)buf = hwcap;
> > res = 0;
> > } else
> > res = ENOENT;
> > break;
> > case AT_HWCAP2:
> > - if (hwcap2_present && buflen == sizeof(u_long)) {
> > + if (hwcap2_present && buflen_ == sizeof(u_long)) {
> > *(u_long *)buf = hwcap2;
> > res = 0;
> > } else
> > res = ENOENT;
> > break;
> > case AT_PAGESIZES:
> > - if (pagesizes != NULL && pagesizes_len >= buflen) {
> > - memcpy(buf, pagesizes, buflen);
> > + if (pagesizes != NULL && pagesizes_len >= buflen_) {
> > + memcpy(buf, pagesizes, buflen_);
> > res = 0;
> > } else
> > res = ENOENT;
> > break;
> > case AT_PAGESZ:
> > - if (buflen == sizeof(int)) {
> > + if (buflen_ == sizeof(int)) {
> > if (pagesize != 0) {
> > *(int *)buf = pagesize;
> > res = 0;
> > @@ -305,7 +311,7 @@ _elf_aux_info(int aux, void *buf, int buflen)
> > res = EINVAL;
> > break;
> > case
svn commit: r366211 - in releng/12.2: sys/fs/fuse tests/sys/fs/fusefs
Author: asomers
Date: Mon Sep 28 00:23:59 2020
New Revision: 366211
URL: https://svnweb.freebsd.org/changeset/base/366211
Log:
MF stable/12 r366190:
fusefs: fix mmap'd writes in direct_io mode
If a FUSE server returns FOPEN_DIRECT_IO in response to FUSE_OPEN, that
instructs the kernel to bypass the page cache for that file. This feature
is also known by libfuse's name: "direct_io".
However, when accessing a file via mmap, there is no possible way to bypass
the cache completely. This change fixes a deadlock that would happen when
an mmap'd write tried to invalidate a portion of the cache, wrongly assuming
that a write couldn't possibly come from cache if direct_io were set.
Arguably, we could instead disable mmap for files with FOPEN_DIRECT_IO set.
But allowing it is less likely to cause user complaints, and is more in
keeping with the spirit of open(2), where O_DIRECT instructs the kernel to
"reduce", not "eliminate" cache effects.
PR: 247276
Approved by: re (gjb)
Reported by: trape...@spawn.link
Reviewed by: cem
Differential Revision:https://reviews.freebsd.org/D26485
Modified:
releng/12.2/sys/fs/fuse/fuse_io.c
releng/12.2/tests/sys/fs/fusefs/write.cc
Directory Properties:
releng/12.2/ (props changed)
Modified: releng/12.2/sys/fs/fuse/fuse_io.c
==
--- releng/12.2/sys/fs/fuse/fuse_io.c Sun Sep 27 23:01:54 2020
(r366210)
+++ releng/12.2/sys/fs/fuse/fuse_io.c Mon Sep 28 00:23:59 2020
(r366211)
@@ -291,6 +291,7 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
fuse_vnode_update(vp, FN_MTIMECHANGE | FN_CTIMECHANGE);
if (directio) {
off_t start, end, filesize;
+ bool pages = (ioflag & IO_VMIO) != 0;
SDT_PROBE2(fusefs, , io, trace, 1,
"direct write of vnode");
@@ -301,15 +302,14 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
start = uio->uio_offset;
end = start + uio->uio_resid;
- KASSERT((ioflag & (IO_VMIO | IO_DIRECT)) !=
- (IO_VMIO | IO_DIRECT),
- ("IO_DIRECT used for a cache flush?"));
- /* Invalidate the write cache when writing directly */
- err = fuse_inval_buf_range(vp, filesize, start, end);
- if (err)
- return (err);
+ if (!pages) {
+ err = fuse_inval_buf_range(vp, filesize, start,
+ end);
+ if (err)
+ return (err);
+ }
err = fuse_write_directbackend(vp, uio, cred, fufh,
- filesize, ioflag, false);
+ filesize, ioflag, pages);
} else {
SDT_PROBE2(fusefs, , io, trace, 1,
"buffered write of vnode");
Modified: releng/12.2/tests/sys/fs/fusefs/write.cc
==
--- releng/12.2/tests/sys/fs/fusefs/write.ccSun Sep 27 23:01:54 2020
(r366210)
+++ releng/12.2/tests/sys/fs/fusefs/write.ccMon Sep 28 00:23:59 2020
(r366211)
@@ -923,6 +923,76 @@ TEST_F(WriteBack, o_direct)
leak(fd);
}
+TEST_F(WriteBack, direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] = "some_file.txt";
+ const char *CONTENTS = "abcdefgh";
+ uint64_t ino = 42;
+ int fd;
+ ssize_t bufsize = strlen(CONTENTS);
+ uint8_t readbuf[bufsize];
+
+ expect_lookup(RELPATH, ino, 0);
+ expect_open(ino, FOPEN_DIRECT_IO, 1);
+ FuseTest::expect_write(ino, 0, bufsize, bufsize, 0, FUSE_WRITE_CACHE,
+ CONTENTS);
+ expect_read(ino, 0, bufsize, bufsize, CONTENTS);
+
+ fd = open(FULLPATH, O_RDWR);
+ EXPECT_LE(0, fd) << strerror(errno);
+
+ ASSERT_EQ(bufsize, write(fd, CONTENTS, bufsize)) << strerror(errno);
+ /* A subsequent read must query the daemon because cache is empty */
+ ASSERT_EQ(0, lseek(fd, 0, SEEK_SET)) << strerror(errno);
+ ASSERT_EQ(0, fcntl(fd, F_SETFL, 0)) << strerror(errno);
+ ASSERT_EQ(bufsize, read(fd, readbuf, bufsize)) << strerror(errno);
+ leak(fd);
+}
+
+/*
+ * mmap should still be possible even if the server used direct_io. Mmap will
+ * still use the cache, though.
+ *
+ * Regression test for bug 247276
+ * https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=247276
+ */
+TEST_F(WriteBack, mmap_direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] =
svn commit: r366207 - head/lib/libc/gen
Author: asomers
Date: Sun Sep 27 22:26:41 2020
New Revision: 366207
URL: https://svnweb.freebsd.org/changeset/base/366207
Log:
Misc compiler warning fixes in lib/libc
Reviewed by: kevans, imp
MFC after:2 weeks
Differential Revision:https://reviews.freebsd.org/D26534
Modified:
head/lib/libc/gen/auxv.c
head/lib/libc/gen/basename_compat.c
head/lib/libc/gen/crypt.c
head/lib/libc/gen/dirname_compat.c
head/lib/libc/gen/fts-compat.c
head/lib/libc/gen/ftw-compat11.c
head/lib/libc/gen/getentropy.c
Modified: head/lib/libc/gen/auxv.c
==
--- head/lib/libc/gen/auxv.cSun Sep 27 21:43:19 2020(r366206)
+++ head/lib/libc/gen/auxv.cSun Sep 27 22:26:41 2020(r366207)
@@ -67,7 +67,8 @@ __init_elf_aux_vector(void)
}
static pthread_once_t aux_once = PTHREAD_ONCE_INIT;
-static int pagesize, osreldate, canary_len, ncpus, pagesizes_len, bsdflags;
+static int pagesize, osreldate, ncpus, bsdflags;
+static size_t canary_len, pagesizes_len;
static int hwcap_present, hwcap2_present;
static char *canary, *pagesizes, *execpath;
static void *ps_strings, *timekeep;
@@ -245,16 +246,21 @@ int
_elf_aux_info(int aux, void *buf, int buflen)
{
int res;
+ size_t buflen_;
__init_elf_aux_vector();
if (__elf_aux_vector == NULL)
return (ENOSYS);
_once(_once, init_aux);
+ if (buflen < 0)
+ return (EINVAL);
+ buflen_ = (size_t)buflen;
+
switch (aux) {
case AT_CANARY:
- if (canary != NULL && canary_len >= buflen) {
- memcpy(buf, canary, buflen);
+ if (canary != NULL && canary_len >= buflen_) {
+ memcpy(buf, canary, buflen_);
memset(canary, 0, canary_len);
canary = NULL;
res = 0;
@@ -267,35 +273,35 @@ _elf_aux_info(int aux, void *buf, int buflen)
else if (buf == NULL)
res = EINVAL;
else {
- if (strlcpy(buf, execpath, buflen) >= buflen)
+ if (strlcpy(buf, execpath, buflen_) >= buflen_)
res = EINVAL;
else
res = 0;
}
break;
case AT_HWCAP:
- if (hwcap_present && buflen == sizeof(u_long)) {
+ if (hwcap_present && buflen_ == sizeof(u_long)) {
*(u_long *)buf = hwcap;
res = 0;
} else
res = ENOENT;
break;
case AT_HWCAP2:
- if (hwcap2_present && buflen == sizeof(u_long)) {
+ if (hwcap2_present && buflen_ == sizeof(u_long)) {
*(u_long *)buf = hwcap2;
res = 0;
} else
res = ENOENT;
break;
case AT_PAGESIZES:
- if (pagesizes != NULL && pagesizes_len >= buflen) {
- memcpy(buf, pagesizes, buflen);
+ if (pagesizes != NULL && pagesizes_len >= buflen_) {
+ memcpy(buf, pagesizes, buflen_);
res = 0;
} else
res = ENOENT;
break;
case AT_PAGESZ:
- if (buflen == sizeof(int)) {
+ if (buflen_ == sizeof(int)) {
if (pagesize != 0) {
*(int *)buf = pagesize;
res = 0;
@@ -305,7 +311,7 @@ _elf_aux_info(int aux, void *buf, int buflen)
res = EINVAL;
break;
case AT_OSRELDATE:
- if (buflen == sizeof(int)) {
+ if (buflen_ == sizeof(int)) {
if (osreldate != 0) {
*(int *)buf = osreldate;
res = 0;
@@ -315,7 +321,7 @@ _elf_aux_info(int aux, void *buf, int buflen)
res = EINVAL;
break;
case AT_NCPUS:
- if (buflen == sizeof(int)) {
+ if (buflen_ == sizeof(int)) {
if (ncpus != 0) {
*(int *)buf = ncpus;
res = 0;
@@ -325,7 +331,7 @@ _elf_aux_info(int aux, void *buf, int buflen)
res = EINVAL;
break;
case AT_TIMEKEEP:
- if (buflen == sizeof(void *)) {
+ if (buflen_ == sizeof(void *)) {
if (timekeep != NULL) {
*(void **)buf = timekeep;
res = 0;
@@ -335,14 +341,14 @@ _elf_aux_info(int aux, void *buf, int buflen)
res = EINVAL;
svn commit: r366190 - in stable/12: sys/fs/fuse tests/sys/fs/fusefs
Author: asomers
Date: Sun Sep 27 02:59:28 2020
New Revision: 366190
URL: https://svnweb.freebsd.org/changeset/base/366190
Log:
MFC r366121:
fusefs: fix mmap'd writes in direct_io mode
If a FUSE server returns FOPEN_DIRECT_IO in response to FUSE_OPEN, that
instructs the kernel to bypass the page cache for that file. This feature
is also known by libfuse's name: "direct_io".
However, when accessing a file via mmap, there is no possible way to bypass
the cache completely. This change fixes a deadlock that would happen when
an mmap'd write tried to invalidate a portion of the cache, wrongly assuming
that a write couldn't possibly come from cache if direct_io were set.
Arguably, we could instead disable mmap for files with FOPEN_DIRECT_IO set.
But allowing it is less likely to cause user complaints, and is more in
keeping with the spirit of open(2), where O_DIRECT instructs the kernel to
"reduce", not "eliminate" cache effects.
PR: 247276
Reported by: trape...@spawn.link
Reviewed by: cem
Differential Revision:https://reviews.freebsd.org/D26485
Modified:
stable/12/sys/fs/fuse/fuse_io.c
stable/12/tests/sys/fs/fusefs/write.cc
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/fs/fuse/fuse_io.c
==
--- stable/12/sys/fs/fuse/fuse_io.c Sat Sep 26 23:05:38 2020
(r366189)
+++ stable/12/sys/fs/fuse/fuse_io.c Sun Sep 27 02:59:28 2020
(r366190)
@@ -291,6 +291,7 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
fuse_vnode_update(vp, FN_MTIMECHANGE | FN_CTIMECHANGE);
if (directio) {
off_t start, end, filesize;
+ bool pages = (ioflag & IO_VMIO) != 0;
SDT_PROBE2(fusefs, , io, trace, 1,
"direct write of vnode");
@@ -301,15 +302,14 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
start = uio->uio_offset;
end = start + uio->uio_resid;
- KASSERT((ioflag & (IO_VMIO | IO_DIRECT)) !=
- (IO_VMIO | IO_DIRECT),
- ("IO_DIRECT used for a cache flush?"));
- /* Invalidate the write cache when writing directly */
- err = fuse_inval_buf_range(vp, filesize, start, end);
- if (err)
- return (err);
+ if (!pages) {
+ err = fuse_inval_buf_range(vp, filesize, start,
+ end);
+ if (err)
+ return (err);
+ }
err = fuse_write_directbackend(vp, uio, cred, fufh,
- filesize, ioflag, false);
+ filesize, ioflag, pages);
} else {
SDT_PROBE2(fusefs, , io, trace, 1,
"buffered write of vnode");
Modified: stable/12/tests/sys/fs/fusefs/write.cc
==
--- stable/12/tests/sys/fs/fusefs/write.cc Sat Sep 26 23:05:38 2020
(r366189)
+++ stable/12/tests/sys/fs/fusefs/write.cc Sun Sep 27 02:59:28 2020
(r366190)
@@ -923,6 +923,76 @@ TEST_F(WriteBack, o_direct)
leak(fd);
}
+TEST_F(WriteBack, direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] = "some_file.txt";
+ const char *CONTENTS = "abcdefgh";
+ uint64_t ino = 42;
+ int fd;
+ ssize_t bufsize = strlen(CONTENTS);
+ uint8_t readbuf[bufsize];
+
+ expect_lookup(RELPATH, ino, 0);
+ expect_open(ino, FOPEN_DIRECT_IO, 1);
+ FuseTest::expect_write(ino, 0, bufsize, bufsize, 0, FUSE_WRITE_CACHE,
+ CONTENTS);
+ expect_read(ino, 0, bufsize, bufsize, CONTENTS);
+
+ fd = open(FULLPATH, O_RDWR);
+ EXPECT_LE(0, fd) << strerror(errno);
+
+ ASSERT_EQ(bufsize, write(fd, CONTENTS, bufsize)) << strerror(errno);
+ /* A subsequent read must query the daemon because cache is empty */
+ ASSERT_EQ(0, lseek(fd, 0, SEEK_SET)) << strerror(errno);
+ ASSERT_EQ(0, fcntl(fd, F_SETFL, 0)) << strerror(errno);
+ ASSERT_EQ(bufsize, read(fd, readbuf, bufsize)) << strerror(errno);
+ leak(fd);
+}
+
+/*
+ * mmap should still be possible even if the server used direct_io. Mmap will
+ * still use the cache, though.
+ *
+ * Regression test for bug 247276
+ * https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=247276
+ */
+TEST_F(WriteBack, mmap_direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] = "some_file.txt";
+ const char *CONTENTS =
svn commit: r366180 - stable/12/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Sat Sep 26 02:50:28 2020
New Revision: 366180
URL: https://svnweb.freebsd.org/changeset/base/366180
Log:
zfs: Fix resuming receive stream to dataset with mounted clone
My fix for bug 248606 (zfs receive: Input/output error accessing dataset
after resuming interrupted receive), r364412, introduced a regression:
attempting to resume a receive into a dataset with a mounted clone would
fail if that clone were in-use. This change reverts r364412 and fixes it in
a better way.
Background:
When ZFS receives a stream, it may decide to unmount and remount the
destination and all of its children. However, ever since resumable
send/receive was implemented, ZFS has skipped the unmount/remount step when
resuming a stream. I don't know why.
That let to bug 248606. When resuming the stream, ZFS didn't unmount and
remount the destination, leaving a destroyed dataset mounted.
My original fix was to always unmount and remount when resuming a receive,
but that caused other problems, like bug 249579. A better solution is to
unmount and remount when resuming a receive of a stream that would've
unmounted and remounted when it was new.
Direct commit to stable/12 because head has moved to OpenZFS. The bug
exists there, too, but a change to the OpenZFS code can't be merged to the
old ZFS code.
PR: 249579
Reviewed by: mmacy
MFC after:1 week
Sponsored by: Axcient
Modified:
stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Modified: stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sat Sep 26 00:58:27 2020(r366179)
+++ stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sat Sep 26 02:50:28 2020(r366180)
@@ -3301,7 +3301,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
boolean_t resuming = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_RESUMING;
stream_wantsnewfs = (drrb->drr_fromguid == 0 ||
- (drrb->drr_flags & DRR_FLAG_CLONE) || originsnap) && !resuming;
+ (drrb->drr_flags & DRR_FLAG_CLONE) || originsnap);
if (stream_wantsnewfs) {
/*
@@ -3433,7 +3433,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
- (stream_wantsnewfs || resuming)) {
+ stream_wantsnewfs) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, 0);
if (clp == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r366132 - head/lib/libc/tests/iconv
Author: asomers
Date: Thu Sep 24 21:42:44 2020
New Revision: 366132
URL: https://svnweb.freebsd.org/changeset/base/366132
Log:
lib/libc/tests/iconv: raise WARNS to 6
MFC after:2 weeks
Modified:
head/lib/libc/tests/iconv/Makefile
head/lib/libc/tests/iconv/iconvctl_test.c
Modified: head/lib/libc/tests/iconv/Makefile
==
--- head/lib/libc/tests/iconv/Makefile Thu Sep 24 21:39:09 2020
(r366131)
+++ head/lib/libc/tests/iconv/Makefile Thu Sep 24 21:42:44 2020
(r366132)
@@ -3,6 +3,5 @@
TESTSDIR= ${TESTSBASE}/lib/libc/iconv
ATF_TESTS_C+= iconvctl_test
-WARNS?=2
.include
Modified: head/lib/libc/tests/iconv/iconvctl_test.c
==
--- head/lib/libc/tests/iconv/iconvctl_test.c Thu Sep 24 21:39:09 2020
(r366131)
+++ head/lib/libc/tests/iconv/iconvctl_test.c Thu Sep 24 21:42:44 2020
(r366132)
@@ -30,7 +30,7 @@ __FBSDID("$FreeBSD$");
#include
-void
+static void
test_trivialp(const char *src, const char *dst, int expected)
{
iconv_t ic;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r366131 - head/lib/libc/tests/sys
Author: asomers
Date: Thu Sep 24 21:39:09 2020
New Revision: 366131
URL: https://svnweb.freebsd.org/changeset/base/366131
Log:
lib/libc/tests/sys: raise WARNS to 6
MFC after:2 weeks
Modified:
head/lib/libc/tests/sys/Makefile
Modified: head/lib/libc/tests/sys/Makefile
==
--- head/lib/libc/tests/sys/MakefileThu Sep 24 20:01:31 2020
(r366130)
+++ head/lib/libc/tests/sys/MakefileThu Sep 24 21:39:09 2020
(r366131)
@@ -78,12 +78,6 @@ LIBADD.timer_create_test+= rt
SRCS.mlock_test+= mlock_helper.c
SRCS.setrlimit_test+= mlock_helper.c
-.if ${COMPILER_TYPE} == "gcc"
-WARNS?=3
-.else
-WARNS?=4
-.endif
-
FILESGROUPS+= truncate_test_FILES
truncate_test_FILES= truncate_test.root_owned
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r366121 - in head: sys/fs/fuse tests/sys/fs/fusefs
Author: asomers
Date: Thu Sep 24 16:27:53 2020
New Revision: 366121
URL: https://svnweb.freebsd.org/changeset/base/366121
Log:
fusefs: fix mmap'd writes in direct_io mode
If a FUSE server returns FOPEN_DIRECT_IO in response to FUSE_OPEN, that
instructs the kernel to bypass the page cache for that file. This feature
is also known by libfuse's name: "direct_io".
However, when accessing a file via mmap, there is no possible way to bypass
the cache completely. This change fixes a deadlock that would happen when
an mmap'd write tried to invalidate a portion of the cache, wrongly assuming
that a write couldn't possibly come from cache if direct_io were set.
Arguably, we could instead disable mmap for files with FOPEN_DIRECT_IO set.
But allowing it is less likely to cause user complaints, and is more in
keeping with the spirit of open(2), where O_DIRECT instructs the kernel to
"reduce", not "eliminate" cache effects.
PR: 247276
Reported by: trape...@spawn.link
Reviewed by: cem
MFC after:3 days
Differential Revision:https://reviews.freebsd.org/D26485
Modified:
head/sys/fs/fuse/fuse_io.c
head/tests/sys/fs/fusefs/write.cc
Modified: head/sys/fs/fuse/fuse_io.c
==
--- head/sys/fs/fuse/fuse_io.c Thu Sep 24 16:21:30 2020(r366120)
+++ head/sys/fs/fuse/fuse_io.c Thu Sep 24 16:27:53 2020(r366121)
@@ -291,6 +291,7 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
fuse_vnode_update(vp, FN_MTIMECHANGE | FN_CTIMECHANGE);
if (directio) {
off_t start, end, filesize;
+ bool pages = (ioflag & IO_VMIO) != 0;
SDT_PROBE2(fusefs, , io, trace, 1,
"direct write of vnode");
@@ -301,15 +302,14 @@ fuse_io_dispatch(struct vnode *vp, struct uio *uio, in
start = uio->uio_offset;
end = start + uio->uio_resid;
- KASSERT((ioflag & (IO_VMIO | IO_DIRECT)) !=
- (IO_VMIO | IO_DIRECT),
- ("IO_DIRECT used for a cache flush?"));
- /* Invalidate the write cache when writing directly */
- err = fuse_inval_buf_range(vp, filesize, start, end);
- if (err)
- return (err);
+ if (!pages) {
+ err = fuse_inval_buf_range(vp, filesize, start,
+ end);
+ if (err)
+ return (err);
+ }
err = fuse_write_directbackend(vp, uio, cred, fufh,
- filesize, ioflag, false);
+ filesize, ioflag, pages);
} else {
SDT_PROBE2(fusefs, , io, trace, 1,
"buffered write of vnode");
Modified: head/tests/sys/fs/fusefs/write.cc
==
--- head/tests/sys/fs/fusefs/write.cc Thu Sep 24 16:21:30 2020
(r366120)
+++ head/tests/sys/fs/fusefs/write.cc Thu Sep 24 16:27:53 2020
(r366121)
@@ -923,6 +923,76 @@ TEST_F(WriteBack, o_direct)
leak(fd);
}
+TEST_F(WriteBack, direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] = "some_file.txt";
+ const char *CONTENTS = "abcdefgh";
+ uint64_t ino = 42;
+ int fd;
+ ssize_t bufsize = strlen(CONTENTS);
+ uint8_t readbuf[bufsize];
+
+ expect_lookup(RELPATH, ino, 0);
+ expect_open(ino, FOPEN_DIRECT_IO, 1);
+ FuseTest::expect_write(ino, 0, bufsize, bufsize, 0, FUSE_WRITE_CACHE,
+ CONTENTS);
+ expect_read(ino, 0, bufsize, bufsize, CONTENTS);
+
+ fd = open(FULLPATH, O_RDWR);
+ EXPECT_LE(0, fd) << strerror(errno);
+
+ ASSERT_EQ(bufsize, write(fd, CONTENTS, bufsize)) << strerror(errno);
+ /* A subsequent read must query the daemon because cache is empty */
+ ASSERT_EQ(0, lseek(fd, 0, SEEK_SET)) << strerror(errno);
+ ASSERT_EQ(0, fcntl(fd, F_SETFL, 0)) << strerror(errno);
+ ASSERT_EQ(bufsize, read(fd, readbuf, bufsize)) << strerror(errno);
+ leak(fd);
+}
+
+/*
+ * mmap should still be possible even if the server used direct_io. Mmap will
+ * still use the cache, though.
+ *
+ * Regression test for bug 247276
+ * https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=247276
+ */
+TEST_F(WriteBack, mmap_direct_io)
+{
+ const char FULLPATH[] = "mountpoint/some_file.txt";
+ const char RELPATH[] = "some_file.txt";
+ const char *CONTENTS = "abcdefgh";
+ uint64_t ino = 42;
+ int fd;
+ size_t len;
+ ssize_t bufsize =
svn commit: r366118 - head/sys/kern
Author: asomers
Date: Thu Sep 24 15:38:01 2020
New Revision: 366118
URL: https://svnweb.freebsd.org/changeset/base/366118
Log:
Fix some signed/unsigned comparison warnings in NFS
Reviewed by: rmacklem
MFC after:2 weeks
Differential Revision:https://reviews.freebsd.org/D26533
Modified:
head/sys/kern/subr_acl_nfs4.c
Modified: head/sys/kern/subr_acl_nfs4.c
==
--- head/sys/kern/subr_acl_nfs4.c Thu Sep 24 15:34:47 2020
(r366117)
+++ head/sys/kern/subr_acl_nfs4.c Thu Sep 24 15:38:01 2020
(r366118)
@@ -342,9 +342,9 @@ _acl_append(struct acl *aclp, acl_tag_t tag, acl_perm_
}
static struct acl_entry *
-_acl_duplicate_entry(struct acl *aclp, int entry_index)
+_acl_duplicate_entry(struct acl *aclp, unsigned entry_index)
{
- int i;
+ unsigned i;
KASSERT(aclp->acl_cnt + 1 <= ACL_MAX_ENTRIES,
("aclp->acl_cnt + 1 <= ACL_MAX_ENTRIES"));
@@ -361,7 +361,8 @@ static void
acl_nfs4_sync_acl_from_mode_draft(struct acl *aclp, mode_t mode,
int file_owner_id)
{
- int i, meets, must_append;
+ int meets, must_append;
+ unsigned i;
struct acl_entry *entry, *copy, *previous,
*a1, *a2, *a3, *a4, *a5, *a6;
mode_t amode;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
Go ahead and commit. Consider it reviewed by me. And if I understand correctly, this commit means there's no need for a special updating procedure, right? On Tue, Sep 22, 2020 at 6:55 PM Warner Losh wrote: > > > On Tue, Sep 22, 2020 at 5:17 PM Kyle Evans wrote: > >> On Tue, Sep 22, 2020, 17:02 Warner Losh wrote: >> >>> >>> >>> On Tue, Sep 22, 2020 at 3:55 PM Kyle Evans wrote: >>> On Tue, Sep 22, 2020 at 4:53 PM Ian Lepore wrote: > > On Tue, 2020-09-22 at 15:50 -0600, Warner Losh wrote: > > I think it's a great leap sideways, but I've done cp /dev/null foo to > > clear > > it out for 35 years now... It's why it feels like a workaround. > > > > Though it is a legit optimization, no matter the feelings. As for > > clearer, > > I'm less sure since then I have to remember what the : operator does. > > > > Warner > > > > For me, :> is idiomatic (but ugly). > > On the other hand, the cp /dev/null had a nice dogfooding aspect to > it... when we broke cp by accident, its use in the build system was the > first alarm to go off. > > --Ian > To be honest, this is a case that really should be covered by regression tests somewhere. >>> >>> It should (but isn't yet). >>> >>> Ian is right for old-school FreeBSD thinking. In that thinking the build >>> system should use an eclectic mix of tools to act as a fire-wall against >>> accidental breakage. >>> >>> Complete, effective, test suites give much better coverage... if they >>> are run... >>> >>> So until we run tests frequently, with loud regression squawking that's >>> as effective as build breakage, I tend to fall in the 'all of the above' >>> camp until that's in place... :) >>> >>> Warner >>> >>> P.S. though not, if I suppose, if it means that we're slowing down the >>> regression coverage uptake... >>> >> >> -- >> >> The test build was fine, please confirm if I can commit it or if someone >> else would like to write the UPDATING notice or start bootstrapping cp on >> systems that were affected. I'm not comfortable with not taking any path at >> all here, but this is a lot of friction for a small mechanical change to >> ease the pain. >> > > Sorry if I wasn't clear: I'm not objecting to the quick mechanical change > so much as complaining that I wish we had better test coverage. Don't let > that stop you from doing what's right (or I can if you'd like). > > Warner > ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
It doesn't feel like a workaround to me. I think Kyle's version is clearer
than the original.
On Tue, Sep 22, 2020 at 3:45 PM Warner Losh wrote:
>
>
> On Tue, Sep 22, 2020 at 3:42 PM Kyle Evans wrote:
>
>> cp is already fixed, people are still feeling the fallout of being
>> within those revisions and needing to bootstrap their own cp. We can
>> reduce the number of components these invocations rely on trivially to
>> shell built-in mechanics, why not do so?
>>
>
> Fair point. I just bristle at putting workarounds in for valid /bin/sh
> syntax because we opposed for a few days. so long as it's an unconditional
> clearing of the file to be zero length, I'm OK with that.
>
> Warner
>
>
>> On Tue, Sep 22, 2020 at 4:40 PM Warner Losh wrote:
>> >
>> > So why do we need a workaround at all? cp /dev/null has been fixed, and
>> that's way more important to get right.
>> >
>> > I don't want to paper-over issues with this at all, though if we use
>> the host's (now broken) cp, I suppose that might be OK in the short term.
>> If that's the case, then maybe this is OK.
>> >
>> > Otherwise, I'd strongly prefer we fix cp.
>> >
>> > Warner
>> >
>> > On Tue, Sep 22, 2020 at 3:31 PM Alan Somers
>> wrote:
>> >>
>> >> +1.
>> >>
>> >> On Tue, Sep 22, 2020 at 3:27 PM Kyle Evans wrote:
>> >>>
>> >>> I'm running a build at the suggestion of mjg to confirm there aren't
>> >>> any others hiding that can be converted, and I will commit after I've
>> >>> verified that this is it.
>> >>>
>> >>> On Tue, Sep 22, 2020 at 4:02 PM Alan Somers
>> wrote:
>> >>> >
>> >>> > LGTM.
>> >>> >
>> >>> > On Tue, Sep 22, 2020 at 2:59 PM Kyle Evans
>> wrote:
>> >>> >>
>> >>> >> Perhaps:
>> >>> >>
>> >>> >> diff --git a/stand/i386/zfsboot/Makefile
>> b/stand/i386/zfsboot/Makefile
>> >>> >> index ff315abc0ef..7e362b43a39 100644
>> >>> >> --- a/stand/i386/zfsboot/Makefile
>> >>> >> +++ b/stand/i386/zfsboot/Makefile
>> >>> >> @@ -81,7 +81,7 @@ zfsboot.ld: zfsboot.ldr zfsboot.bin ${BTXKERN}
>> >>> >> -o ${.TARGET} -P 1 zfsboot.bin
>> >>> >>
>> >>> >> zfsboot.ldr:
>> >>> >> - cp /dev/null ${.TARGET}
>> >>> >> + :> ${.TARGET}
>> >>> >>
>> >>> >> zfsboot.bin: zfsboot.out
>> >>> >> ${OBJCOPY} -S -O binary zfsboot.out ${.TARGET}
>> >>> >> diff --git a/stand/libsa/Makefile b/stand/libsa/Makefile
>> >>> >> index effece9e01b..63cd46a9c54 100644
>> >>> >> --- a/stand/libsa/Makefile
>> >>> >> +++ b/stand/libsa/Makefile
>> >>> >> @@ -122,7 +122,7 @@ beforedepend:
>> >>> >> ln -sf ${SRCTOP}/include/arpa/inet.h arpa/inet.h; \
>> >>> >> ln -sf ${SRCTOP}/include/arpa/tftp.h arpa/tftp.h; \
>> >>> >> for i in _time.h _strings.h _string.h; do \
>> >>> >> - [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
>> >>> >> + [ -f xlocale/$$i ] || :> xlocale/$$i; \
>> >>> >> done; \
>> >>> >> for i in ${STAND_H_INC}; do \
>> >>> >> ln -sf ${SASRC}/stand.h $$i; \
>> >>> >>
>> >>> >>
>> >>> >> On Tue, Sep 22, 2020 at 3:58 PM Alan Somers
>> wrote:
>> >>> >> >
>> >>> >> > Looks like two places in stand. Is there any reason why
>> Mateusz's suggestion wouldn't work?
>> >>> >> >
>> >>> >> > > rg -g Makefile 'cp.*/dev/null'
>> >>> >> > stand/libsa/Makefile
>> >>> >> > 125: [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
>> >>> >> >
>> >>> >> > stand/i386/zfsboot/Makefile
>> >>> >> > 82: cp /dev/null ${.TARGET}
>> >>> >> >
>> >>> >> > On Tue, Sep 22, 2020 at 2:54 PM Mateusz Guzik
>> wrote:
>> >>> >> >>
>> >>> >> >> Can we instead add a workaround
Re: svn commit: r365643 - head/bin/cp
+1.
On Tue, Sep 22, 2020 at 3:27 PM Kyle Evans wrote:
> I'm running a build at the suggestion of mjg to confirm there aren't
> any others hiding that can be converted, and I will commit after I've
> verified that this is it.
>
> On Tue, Sep 22, 2020 at 4:02 PM Alan Somers wrote:
> >
> > LGTM.
> >
> > On Tue, Sep 22, 2020 at 2:59 PM Kyle Evans wrote:
> >>
> >> Perhaps:
> >>
> >> diff --git a/stand/i386/zfsboot/Makefile b/stand/i386/zfsboot/Makefile
> >> index ff315abc0ef..7e362b43a39 100644
> >> --- a/stand/i386/zfsboot/Makefile
> >> +++ b/stand/i386/zfsboot/Makefile
> >> @@ -81,7 +81,7 @@ zfsboot.ld: zfsboot.ldr zfsboot.bin ${BTXKERN}
> >> -o ${.TARGET} -P 1 zfsboot.bin
> >>
> >> zfsboot.ldr:
> >> - cp /dev/null ${.TARGET}
> >> + :> ${.TARGET}
> >>
> >> zfsboot.bin: zfsboot.out
> >> ${OBJCOPY} -S -O binary zfsboot.out ${.TARGET}
> >> diff --git a/stand/libsa/Makefile b/stand/libsa/Makefile
> >> index effece9e01b..63cd46a9c54 100644
> >> --- a/stand/libsa/Makefile
> >> +++ b/stand/libsa/Makefile
> >> @@ -122,7 +122,7 @@ beforedepend:
> >> ln -sf ${SRCTOP}/include/arpa/inet.h arpa/inet.h; \
> >> ln -sf ${SRCTOP}/include/arpa/tftp.h arpa/tftp.h; \
> >> for i in _time.h _strings.h _string.h; do \
> >> - [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
> >> + [ -f xlocale/$$i ] || :> xlocale/$$i; \
> >> done; \
> >> for i in ${STAND_H_INC}; do \
> >> ln -sf ${SASRC}/stand.h $$i; \
> >>
> >>
> >> On Tue, Sep 22, 2020 at 3:58 PM Alan Somers
> wrote:
> >> >
> >> > Looks like two places in stand. Is there any reason why Mateusz's
> suggestion wouldn't work?
> >> >
> >> > > rg -g Makefile 'cp.*/dev/null'
> >> > stand/libsa/Makefile
> >> > 125: [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
> >> >
> >> > stand/i386/zfsboot/Makefile
> >> > 82: cp /dev/null ${.TARGET}
> >> >
> >> > On Tue, Sep 22, 2020 at 2:54 PM Mateusz Guzik
> wrote:
> >> >>
> >> >> Can we instead add a workaround to the build tree?
> >> >>
> >> >> Where is cp /dev/null coming from anyway? Perhaps this can be patched
> >> >> to touch the target file.
> >> >>
> >> >> On 9/22/20, Alan Somers wrote:
> >> >> > On Tue, Sep 22, 2020 at 2:48 PM Kyle Evans
> wrote:
> >> >> >
> >> >> >> On Fri, Sep 11, 2020 at 3:49 PM Alan Somers
> wrote:
> >> >> >> >
> >> >> >> > Author: asomers
> >> >> >> > Date: Fri Sep 11 20:49:36 2020
> >> >> >> > New Revision: 365643
> >> >> >> > URL: https://svnweb.freebsd.org/changeset/base/365643
> >> >> >> >
> >> >> >> > Log:
> >> >> >> > cp: fall back to read/write if copy_file_range fails
> >> >> >> >
> >> >> >> > Even though copy_file_range has a file-system agnostic
> version, it
> >> >> >> still
> >> >> >> > fails on devfs (perhaps because the file descriptor is
> non-seekable?)
> >> >> >> In
> >> >> >> > that case, fallback to old-fashioned read/write. Fixes
> >> >> >> > "cp /dev/null /tmp/null"
> >> >> >> >
> >> >> >>
> >> >> >> Hi,
> >> >> >>
> >> >> >> Any objection to adding a quick UPDATING entry for this? I'm
> seeing
> >> >> >> occasional reports of this breakage as recent as today on IRC from
> >> >> >> folks that were a little bit thrown off by this because it throws
> up
> >> >> >> fairly far into the build and looks like a stand build regression
> >> >> >> instead of a cp regression.
> >> >> >>
> >> >> >> Thanks,
> >> >> >>
> >> >> >> Kyle Evans
> >> >> >>
> >> >> >
> >> >> > No objection. Can you suggest the proper wording?
> >> >> > ___
> >> >> > svn-src-all@freebsd.org mailing list
> >> >> > https://lists.freebsd.org/mailman/listinfo/svn-src-all
> >> >> > To unsubscribe, send any mail to "
> svn-src-all-unsubscr...@freebsd.org"
> >> >> >
> >> >>
> >> >>
> >> >> --
> >> >> Mateusz Guzik
>
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
LGTM.
On Tue, Sep 22, 2020 at 2:59 PM Kyle Evans wrote:
> Perhaps:
>
> diff --git a/stand/i386/zfsboot/Makefile b/stand/i386/zfsboot/Makefile
> index ff315abc0ef..7e362b43a39 100644
> --- a/stand/i386/zfsboot/Makefile
> +++ b/stand/i386/zfsboot/Makefile
> @@ -81,7 +81,7 @@ zfsboot.ld: zfsboot.ldr zfsboot.bin ${BTXKERN}
> -o ${.TARGET} -P 1 zfsboot.bin
>
> zfsboot.ldr:
> - cp /dev/null ${.TARGET}
> + :> ${.TARGET}
>
> zfsboot.bin: zfsboot.out
> ${OBJCOPY} -S -O binary zfsboot.out ${.TARGET}
> diff --git a/stand/libsa/Makefile b/stand/libsa/Makefile
> index effece9e01b..63cd46a9c54 100644
> --- a/stand/libsa/Makefile
> +++ b/stand/libsa/Makefile
> @@ -122,7 +122,7 @@ beforedepend:
> ln -sf ${SRCTOP}/include/arpa/inet.h arpa/inet.h; \
> ln -sf ${SRCTOP}/include/arpa/tftp.h arpa/tftp.h; \
> for i in _time.h _strings.h _string.h; do \
> - [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
> + [ -f xlocale/$$i ] || :> xlocale/$$i; \
> done; \
> for i in ${STAND_H_INC}; do \
> ln -sf ${SASRC}/stand.h $$i; \
>
>
> On Tue, Sep 22, 2020 at 3:58 PM Alan Somers wrote:
> >
> > Looks like two places in stand. Is there any reason why Mateusz's
> suggestion wouldn't work?
> >
> > > rg -g Makefile 'cp.*/dev/null'
> > stand/libsa/Makefile
> > 125: [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
> >
> > stand/i386/zfsboot/Makefile
> > 82: cp /dev/null ${.TARGET}
> >
> > On Tue, Sep 22, 2020 at 2:54 PM Mateusz Guzik wrote:
> >>
> >> Can we instead add a workaround to the build tree?
> >>
> >> Where is cp /dev/null coming from anyway? Perhaps this can be patched
> >> to touch the target file.
> >>
> >> On 9/22/20, Alan Somers wrote:
> >> > On Tue, Sep 22, 2020 at 2:48 PM Kyle Evans
> wrote:
> >> >
> >> >> On Fri, Sep 11, 2020 at 3:49 PM Alan Somers
> wrote:
> >> >> >
> >> >> > Author: asomers
> >> >> > Date: Fri Sep 11 20:49:36 2020
> >> >> > New Revision: 365643
> >> >> > URL: https://svnweb.freebsd.org/changeset/base/365643
> >> >> >
> >> >> > Log:
> >> >> > cp: fall back to read/write if copy_file_range fails
> >> >> >
> >> >> > Even though copy_file_range has a file-system agnostic version,
> it
> >> >> still
> >> >> > fails on devfs (perhaps because the file descriptor is
> non-seekable?)
> >> >> In
> >> >> > that case, fallback to old-fashioned read/write. Fixes
> >> >> > "cp /dev/null /tmp/null"
> >> >> >
> >> >>
> >> >> Hi,
> >> >>
> >> >> Any objection to adding a quick UPDATING entry for this? I'm seeing
> >> >> occasional reports of this breakage as recent as today on IRC from
> >> >> folks that were a little bit thrown off by this because it throws up
> >> >> fairly far into the build and looks like a stand build regression
> >> >> instead of a cp regression.
> >> >>
> >> >> Thanks,
> >> >>
> >> >> Kyle Evans
> >> >>
> >> >
> >> > No objection. Can you suggest the proper wording?
> >> > ___
> >> > svn-src-all@freebsd.org mailing list
> >> > https://lists.freebsd.org/mailman/listinfo/svn-src-all
> >> > To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org
> "
> >> >
> >>
> >>
> >> --
> >> Mateusz Guzik
>
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
Looks like two places in stand. Is there any reason why Mateusz's
suggestion wouldn't work?
> rg -g Makefile 'cp.*/dev/null'
stand/libsa/Makefile
125: [ -f xlocale/$$i ] || cp /dev/null xlocale/$$i; \
stand/i386/zfsboot/Makefile
82: cp /dev/null ${.TARGET}
On Tue, Sep 22, 2020 at 2:54 PM Mateusz Guzik wrote:
> Can we instead add a workaround to the build tree?
>
> Where is cp /dev/null coming from anyway? Perhaps this can be patched
> to touch the target file.
>
> On 9/22/20, Alan Somers wrote:
> > On Tue, Sep 22, 2020 at 2:48 PM Kyle Evans wrote:
> >
> >> On Fri, Sep 11, 2020 at 3:49 PM Alan Somers
> wrote:
> >> >
> >> > Author: asomers
> >> > Date: Fri Sep 11 20:49:36 2020
> >> > New Revision: 365643
> >> > URL: https://svnweb.freebsd.org/changeset/base/365643
> >> >
> >> > Log:
> >> > cp: fall back to read/write if copy_file_range fails
> >> >
> >> > Even though copy_file_range has a file-system agnostic version, it
> >> still
> >> > fails on devfs (perhaps because the file descriptor is
> non-seekable?)
> >> In
> >> > that case, fallback to old-fashioned read/write. Fixes
> >> > "cp /dev/null /tmp/null"
> >> >
> >>
> >> Hi,
> >>
> >> Any objection to adding a quick UPDATING entry for this? I'm seeing
> >> occasional reports of this breakage as recent as today on IRC from
> >> folks that were a little bit thrown off by this because it throws up
> >> fairly far into the build and looks like a stand build regression
> >> instead of a cp regression.
> >>
> >> Thanks,
> >>
> >> Kyle Evans
> >>
> >
> > No objection. Can you suggest the proper wording?
> > ___
> > svn-src-all@freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/svn-src-all
> > To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
> >
>
>
> --
> Mateusz Guzik
>
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
On Tue, Sep 22, 2020 at 2:48 PM Kyle Evans wrote: > On Fri, Sep 11, 2020 at 3:49 PM Alan Somers wrote: > > > > Author: asomers > > Date: Fri Sep 11 20:49:36 2020 > > New Revision: 365643 > > URL: https://svnweb.freebsd.org/changeset/base/365643 > > > > Log: > > cp: fall back to read/write if copy_file_range fails > > > > Even though copy_file_range has a file-system agnostic version, it > still > > fails on devfs (perhaps because the file descriptor is non-seekable?) > In > > that case, fallback to old-fashioned read/write. Fixes > > "cp /dev/null /tmp/null" > > > > Hi, > > Any objection to adding a quick UPDATING entry for this? I'm seeing > occasional reports of this breakage as recent as today on IRC from > folks that were a little bit thrown off by this because it throws up > fairly far into the build and looks like a stand build regression > instead of a cp regression. > > Thanks, > > Kyle Evans > No objection. Can you suggest the proper wording? ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365956 - head/tools/regression/fsx
Author: asomers
Date: Mon Sep 21 17:48:28 2020
New Revision: 365956
URL: https://svnweb.freebsd.org/changeset/base/365956
Log:
fsx: fix build with WARNS=6
* signed/unsigned comparisons
* use standard warn(3)
* Suppress warnings about local vars and funcs not declared static
* const-correctness
* declaration shadows a variable in the global scope
Reviewed by: kevans
MFC after:2 weeks
Differential Revision:https://reviews.freebsd.org/D26516
Modified:
head/tools/regression/fsx/Makefile
head/tools/regression/fsx/fsx.c
Modified: head/tools/regression/fsx/Makefile
==
--- head/tools/regression/fsx/Makefile Mon Sep 21 17:28:41 2020
(r365955)
+++ head/tools/regression/fsx/Makefile Mon Sep 21 17:48:28 2020
(r365956)
@@ -4,4 +4,9 @@ PROG= fsx
MAN=
+NO_WMISSING_VARIABLE_DECLARATIONS=
+
.include
+
+# Don't require static declarations. It's line noise in a single-file program.
+CWARNFLAGS+= -Wno-strict-prototypes -Wno-missing-prototypes
Modified: head/tools/regression/fsx/fsx.c
==
--- head/tools/regression/fsx/fsx.c Mon Sep 21 17:28:41 2020
(r365955)
+++ head/tools/regression/fsx/fsx.c Mon Sep 21 17:48:28 2020
(r365956)
@@ -112,7 +112,7 @@ int closeprob = 0; /* -c flag */
intinvlprob = 0; /* -i flag */
intdebug = 0; /* -d flag */
unsigned long debugstart = 0; /* -D flag */
-unsigned long maxfilelen = 256 * 1024;/* -l flag */
+off_t maxfilelen = 256 * 1024;/* -l flag */
intsizechecks = 1; /* -n flag disables them */
intmaxoplen = 64 * 1024; /* -o flag */
intquiet = 0; /* -q flag */
@@ -138,33 +138,8 @@ int invl = 0;
void
-vwarnc(code, fmt, ap)
- int code;
- const char *fmt;
- va_list ap;
+prt(const char *fmt, ...)
{
- fprintf(stderr, "fsx: ");
- if (fmt != NULL) {
- vfprintf(stderr, fmt, ap);
- fprintf(stderr, ": ");
- }
- fprintf(stderr, "%s\n", strerror(code));
-}
-
-
-void
-warn(const char * fmt, ...)
-{
- va_list ap;
- va_start(ap, fmt);
- vwarnc(errno, fmt, ap);
- va_end(ap);
-}
-
-
-void
-prt(char *fmt, ...)
-{
va_list args;
va_start(args, fmt);
@@ -179,7 +154,7 @@ prt(char *fmt, ...)
}
void
-prterr(char *prefix)
+prterr(const char *prefix)
{
prt("%s%s%s\n", prefix, prefix ? ": " : "", strerror(errno));
}
@@ -317,12 +292,12 @@ logdump(void)
void
-save_buffer(char *buffer, off_t bufferlength, int fd)
+save_buffer(char *buffer, off_t bufferlength, int savefd)
{
off_t ret;
ssize_t byteswritten;
- if (fd <= 0 || bufferlength == 0)
+ if (savefd <= 0 || bufferlength == 0)
return;
if (bufferlength > SSIZE_MAX) {
@@ -330,7 +305,7 @@ save_buffer(char *buffer, off_t bufferlength, int fd)
exit(67);
}
if (lite) {
- off_t size_by_seek = lseek(fd, (off_t)0, SEEK_END);
+ off_t size_by_seek = lseek(savefd, (off_t)0, SEEK_END);
if (size_by_seek == (off_t)-1)
prterr("save_buffer: lseek eof");
else if (bufferlength > size_by_seek) {
@@ -340,11 +315,11 @@ save_buffer(char *buffer, off_t bufferlength, int fd)
}
}
- ret = lseek(fd, (off_t)0, SEEK_SET);
+ ret = lseek(savefd, (off_t)0, SEEK_SET);
if (ret == (off_t)-1)
prterr("save_buffer: lseek 0");
- byteswritten = write(fd, buffer, (size_t)bufferlength);
+ byteswritten = write(savefd, buffer, (size_t)bufferlength);
if (byteswritten != bufferlength) {
if (byteswritten == -1)
prterr("save_buffer write");
@@ -458,10 +433,10 @@ check_trunc_hack(void)
void
-doread(unsigned offset, unsigned size)
+doread(off_t offset, off_t size)
{
off_t ret;
- unsigned iret;
+ ssize_t iret;
offset -= offset % readbdy;
if (size == 0) {
@@ -509,7 +484,7 @@ doread(unsigned offset, unsigned size)
void
-check_eofpage(char *s, unsigned offset, char *p, int size)
+check_eofpage(const char *s, unsigned offset, char *p, int size)
{
uintptr_t last_page, should_be_zero;
@@ -592,7 +567,7 @@ domapread(unsigned offset, unsigned size)
void
-gendata(char *original_buf, char *good_buf, unsigned offset, unsigned size)
+gendata(unsigned offset, unsigned size)
{
while (size--) {
good_buf[offset] = testcalls % 256;
@@ -607,7 +582,7 @@ void
dowrite(unsigned offset, unsigned size)
{
off_t ret;
- unsigned iret;
+ ssize_t iret;
offset -= offset % writebdy;
if (size == 0) {
Re: svn commit: r365643 - head/bin/cp
On Sat, Sep 19, 2020 at 5:32 PM Konstantin Belousov wrote: > On Sat, Sep 19, 2020 at 11:18:56PM +, Rick Macklem wrote: > > Alan Somers wrote: > > >On Fri, Sep 11, 2020 at 3:52 PM Rick Macklem <mailto:rmack...@uoguelph.ca>> wrote: > > >Konstantin Belousov wrote: > > >>On Fri, Sep 11, 2020 at 08:49:36PM +, Alan Somers wrote: > > >>> Author: asomers > > >>> Date: Fri Sep 11 20:49:36 2020 > > >>> New Revision: 365643 > > >>> URL: https://svnweb.freebsd.org/changeset/base/365643 > > >>> > > >>> Log: > > >>> cp: fall back to read/write if copy_file_range fails > > >>> > > >>> Even though copy_file_range has a file-system agnostic version, it > still > > >>> fails on devfs (perhaps because the file descriptor is > non-seekable?) In > > >>> that case, fallback to old-fashioned read/write. Fixes > > >>> "cp /dev/null /tmp/null" > > >> > > >>Devices are seekable. > > >> > > >>The reason for EINVAL is that vn_copy_file_range() checks that both in > and out > > >>vnodes are VREG. For devfs, they are VCHR. > > > > > >I coded the syscall to the Linux man page, which states that EINVAL is > returned > > >if either fd does not refer to a regular file. > > >Having said that, I do not recall testing the VCHR case under Linux. > (ie. It might > > >actually work and the man page turns out to be incorrect?) > > > > > >I will test this case under Linux when I get home next week, rick > > I'll admit I haven't tested this in Linux to see if they do return > EINVAL. > > > > >Since there's no standard, I think it's fine for us to support devfs if > possible. > > 1 - I think this is a good question for a mailing list like > freebsd-current@. > > 2 - I see Linux as the de-facto standard these days and consider POSIX no > > longer relevant, but that's just mho. > > 3 - For NFSv4.2, the Copy operation will fail for non-regular files, so > if you > > do this, you will need to handle the fall-back to using the > generic code. > > (Should be doable, but you need to be aware of this case.) > > > > Having said the above, it is up to the "collective" and not me and, as > such, > > I suggest #1, to see whether others think doing a non-Linux compatible > > version makes sense for FreeBSD? > > I believe that allowing devfs nodes for vn_copy_file() is not very good > idea. For /dev/null driver returns EOF, but think about real devices or > even better, /dev/zero that never EOF its output. > > Is vn_copy_file() interruptible ? I think not. So if insane range is > specified, we have unstoppable copier that fills the disk (at best). > I can think of good use cases for copy_file_range on a device: 1) Network block devices. I don't know if the iSCSI, NBD, or Ceph RBD protocols currently support server-side copies, but it's reasonable that they might. If they ever do, FreeBSD would need copy_file_range to take advantage. 2) CUSE. I think Linux's CUSE already supports copy_file_range, since a CUSE device on Linux is basically just a single-file FUSE file system. We might add support to our CUSE driver someday. 3) zvols. This is the use case that matters the most to me. I have a large amount of data stored in plain files that I would like to convert to zvols. dd should be able to do that using copy_file_range. In my opinion, the utility of those cases outweighs the risk of a long-running interruptible syscall. And in any case, it is documented that copy_file_range may return EINTR. -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365911 - stable/12/share/man/man5
Author: asomers Date: Sat Sep 19 19:48:15 2020 New Revision: 365911 URL: https://svnweb.freebsd.org/changeset/base/365911 Log: MFC r365391, r365415 r365391: nsswitch.conf(5): recommend placing cache after files When cache precedes files, and nscd is configured to allow negative caching, commands like "pw groupadd" can fail. The sequence of events looks like: 1. A command like pkg(8) looks up the group, and finds it absent. 2. pkg invokes pw(8) to add the group 3. pkg queries the group, but nscd says it doesn't exist, since it has a negative cache entry for that group. See also: https://lists.freebsd.org/pipermail/freebsd-current/2012-January/031595.html Reviewed by: bcr (manpages) Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D26184 r365415: nsswitch.conf.5: style fixes Fix some whitespace, and remove the .Tn macro Reported by: mandoc, igor Reviewed by: bcr (manpages) Differential Revision:https://reviews.freebsd.org/D26345 Modified: stable/12/share/man/man5/nsswitch.conf.5 Directory Properties: stable/12/ (props changed) Modified: stable/12/share/man/man5/nsswitch.conf.5 == --- stable/12/share/man/man5/nsswitch.conf.5Sat Sep 19 19:08:27 2020 (r365910) +++ stable/12/share/man/man5/nsswitch.conf.5Sat Sep 19 19:48:15 2020 (r365911) @@ -1,4 +1,4 @@ -.\"$NetBSD: nsswitch.conf.5,v 1.14 1999/03/17 20:19:47 garbled Exp $ +.\" $NetBSD: nsswitch.conf.5,v 1.14 1999/03/17 20:19:47 garbled Exp $ .\" .\" Copyright (c) 1997, 1998, 1999 The NetBSD Foundation, Inc. .\" All rights reserved. @@ -16,7 +16,7 @@ .\"documentation and/or other materials provided with the distribution. .\" 3. All advertising materials mentioning features or use of this software .\"must display the following acknowledgement: -.\"This product includes software developed by Luke Mewburn. +.\"This product includes software developed by Luke Mewburn. .\" 4. The name of the author may not be used to endorse or promote products .\"derived from this software without specific prior written permission. .\" @@ -33,7 +33,7 @@ .\" .\" $FreeBSD$ .\" -.Dd June 6, 2016 +.Dd September 6, 2020 .Dt NSSWITCH.CONF 5 .Os .Sh NAME @@ -48,8 +48,7 @@ file specifies how the .Pp The configuration file controls how a process looks up various databases containing information regarding hosts, users (passwords), groups, etc. -Each database comes from a source (such as local files, DNS, -.Tn NIS , +Each database comes from a source (such as local files, DNS, NIS , and cache), and the order to look up the sources is specified in .Nm . .Pp @@ -180,9 +179,7 @@ Try the next source Return with the current result .El .Ss Format of file -A -.Tn BNF -description of the syntax of +A BNF description of the syntax of .Nm is: .Pp @@ -222,20 +219,24 @@ and continue on anything else (i.e, .Ss Cache You can enable caching for the particular database by specifying .Dq cache -as the first source in the +in the .Nm file. +It should come after +.Dq files , +but before remote sources like +.Dq nis . You should also enable caching for this database in .Xr nscd.conf 5 . -If for the particular query +If for a particular query .Dq cache -source returns success, no further sources are queried. +source returns success, then no further sources are queried. On the other hand, if there are no previously cached data, the query result will be placed into the cache right after all other sources are processed. -Note, that +Note that .Dq cache -requires +requires the .Xr nscd 8 daemon to be running. .Ss Compat mode: +/- syntax @@ -244,12 +245,10 @@ In historical multi-source implementations, the and .Sq - characters are used to specify the importing of user password and -group information from -.Tn NIS . +group information from NIS . Although .Nm -provides alternative methods of accessing distributed sources such as -.Tn NIS , +provides alternative methods of accessing distributed sources such as NIS , specifying a sole source of .Dq compat will provide the historical behaviour. @@ -319,15 +318,14 @@ resides in .Pa /etc . .El .Sh EXAMPLES -To lookup hosts in cache, then in +To lookup hosts in .Pa /etc/hosts -and then from the DNS, and lookup user information from -.Tn NIS -then files, use: +, then in cache, +and then from the DNS, and lookup user information from NIS then files, use: .Pp .Bl -tag -width passwd: -compact .It hosts: -cache files dns +files cache dns .It passwd: nis [notfound=return] files .It group: @@ -349,9 +347,7 @@ entries. .Fx Ns 's .Lb libc provides stubs for compatibility with NSS modules -written for the -.Tn GNU -C Library +written for the GNU C Library .Nm nsswitch interface. However, these stubs only support the use of the @@ -377,10 +373,8 @@ Project, where it appeared first in
svn commit: r365910 - head/lib/libc/gen
Author: asomers
Date: Sat Sep 19 19:08:27 2020
New Revision: 365910
URL: https://svnweb.freebsd.org/changeset/base/365910
Log:
fix integer underflow in getgrnam_r and getpwnam_r
Sometimes nscd(8) will return a 1-byte buffer for a nonexistent entry. This
triggered an integer underflow in grp_unmarshal_func, causing getgrnam_r to
return ERANGE instead of 0.
Fix the user's buffer size check, and add a correct check for a too-small
nscd buffer.
PR: 248932
Event:September 2020 Bugathon
Reviewed by: markj
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision: https://reviews.freebsd.org/D26204
Modified:
head/lib/libc/gen/getgrent.c
head/lib/libc/gen/getpwent.c
Modified: head/lib/libc/gen/getgrent.c
==
--- head/lib/libc/gen/getgrent.cSat Sep 19 18:25:52 2020
(r365909)
+++ head/lib/libc/gen/getgrent.cSat Sep 19 19:08:27 2020
(r365910)
@@ -334,14 +334,27 @@ grp_unmarshal_func(char *buffer, size_t buffer_size, v
orig_buf_size = va_arg(ap, size_t);
ret_errno = va_arg(ap, int *);
- if (orig_buf_size <
- buffer_size - sizeof(struct group) - sizeof(char *)) {
+ if (orig_buf_size + sizeof(struct group) + sizeof(char *) < buffer_size)
+ {
*ret_errno = ERANGE;
return (NS_RETURN);
+ } else if (buffer_size < sizeof(struct group) + sizeof(char *)) {
+ /*
+* nscd(8) sometimes returns buffer_size=1 for nonexistent
+* entries.
+*/
+ *ret_errno = 0;
+ return (NS_NOTFOUND);
}
memcpy(grp, buffer, sizeof(struct group));
memcpy(, buffer + sizeof(struct group), sizeof(char *));
+
+ if (orig_buf_size + sizeof(struct group) + sizeof(char *) +
+ _ALIGN(p) - (size_t)p < buffer_size) {
+ *ret_errno = ERANGE;
+ return (NS_RETURN);
+ }
orig_buf = (char *)_ALIGN(orig_buf);
memcpy(orig_buf, buffer + sizeof(struct group) + sizeof(char *) +
Modified: head/lib/libc/gen/getpwent.c
==
--- head/lib/libc/gen/getpwent.cSat Sep 19 18:25:52 2020
(r365909)
+++ head/lib/libc/gen/getpwent.cSat Sep 19 19:08:27 2020
(r365910)
@@ -389,10 +389,17 @@ pwd_unmarshal_func(char *buffer, size_t buffer_size, v
orig_buf_size = va_arg(ap, size_t);
ret_errno = va_arg(ap, int *);
- if (orig_buf_size <
- buffer_size - sizeof(struct passwd) - sizeof(char *)) {
+ if (orig_buf_size + sizeof(struct passwd) + sizeof(char *) <
+ buffer_size) {
*ret_errno = ERANGE;
return (NS_RETURN);
+ } else if (buffer_size < sizeof(struct passwd) + sizeof(char *)) {
+ /*
+* nscd(8) sometimes returns buffer_size=1 for nonexistent
+* entries.
+*/
+ *ret_errno = 0;
+ return (NS_NOTFOUND);
}
memcpy(pwd, buffer, sizeof(struct passwd));
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365909 - stable/12/tests/sys/opencrypto
Author: asomers
Date: Sat Sep 19 18:25:52 2020
New Revision: 365909
URL: https://svnweb.freebsd.org/changeset/base/365909
Log:
MFC r363361:
tests/sys/opencrypto: use python3
python2 will be EOL soon
Reviewed by: lwhsu, jmg
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25682
Modified:
stable/12/tests/sys/opencrypto/Makefile
stable/12/tests/sys/opencrypto/cryptodev.py
stable/12/tests/sys/opencrypto/cryptotest.py
stable/12/tests/sys/opencrypto/runtests.sh
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/tests/sys/opencrypto/Makefile
==
--- stable/12/tests/sys/opencrypto/Makefile Sat Sep 19 18:02:55 2020
(r365908)
+++ stable/12/tests/sys/opencrypto/Makefile Sat Sep 19 18:25:52 2020
(r365909)
@@ -14,7 +14,7 @@ ATF_TESTS_C+= blake2_test poly1305_test
TAP_TESTS_SH+= runtests
-TEST_METADATA.runtests+= required_programs="python2"
+TEST_METADATA.runtests+= required_programs="python3"
TEST_METADATA.runtests+= required_user="root"
PYMODULES= cryptodev.py cryptodevh.py cryptotest.py
Modified: stable/12/tests/sys/opencrypto/cryptodev.py
==
--- stable/12/tests/sys/opencrypto/cryptodev.py Sat Sep 19 18:02:55 2020
(r365908)
+++ stable/12/tests/sys/opencrypto/cryptodev.py Sat Sep 19 18:25:52 2020
(r365909)
@@ -1,4 +1,4 @@
-#!/usr/local/bin/python2
+#!/usr/local/bin/python3
#
# Copyright (c) 2014 The FreeBSD Foundation
# Copyright 2014 John-Mark Gurney
@@ -30,7 +30,7 @@
# $FreeBSD$
#
-from __future__ import print_function
+
import array
import dpkt
from fcntl import ioctl
Modified: stable/12/tests/sys/opencrypto/cryptotest.py
==
--- stable/12/tests/sys/opencrypto/cryptotest.pySat Sep 19 18:02:55
2020(r365908)
+++ stable/12/tests/sys/opencrypto/cryptotest.pySat Sep 19 18:25:52
2020(r365909)
@@ -1,4 +1,4 @@
-#!/usr/local/bin/python2
+#!/usr/local/bin/python3
#
# Copyright (c) 2014 The FreeBSD Foundation
# All rights reserved.
@@ -29,7 +29,9 @@
# $FreeBSD$
#
-from __future__ import print_function
+
+
+import binascii
import errno
import cryptodev
import itertools
Modified: stable/12/tests/sys/opencrypto/runtests.sh
==
--- stable/12/tests/sys/opencrypto/runtests.sh Sat Sep 19 18:02:55 2020
(r365908)
+++ stable/12/tests/sys/opencrypto/runtests.sh Sat Sep 19 18:25:52 2020
(r365909)
@@ -29,7 +29,7 @@
# $FreeBSD$
#
-: ${PYTHON=python2}
+: ${PYTHON=python3}
if [ ! -d /usr/local/share/nist-kat ]; then
echo "1..0 # SKIP: nist-kat package not installed for test vectors"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365643 - head/bin/cp
On Fri, Sep 11, 2020 at 3:52 PM Rick Macklem wrote: > Konstantin Belousov wrote: > >On Fri, Sep 11, 2020 at 08:49:36PM +0000, Alan Somers wrote: > >> Author: asomers > >> Date: Fri Sep 11 20:49:36 2020 > >> New Revision: 365643 > >> URL: https://svnweb.freebsd.org/changeset/base/365643 > >> > >> Log: > >> cp: fall back to read/write if copy_file_range fails > >> > >> Even though copy_file_range has a file-system agnostic version, it > still > >> fails on devfs (perhaps because the file descriptor is non-seekable?) > In > >> that case, fallback to old-fashioned read/write. Fixes > >> "cp /dev/null /tmp/null" > > > >Devices are seekable. > > > >The reason for EINVAL is that vn_copy_file_range() checks that both in > and out > >vnodes are VREG. For devfs, they are VCHR. > > I coded the syscall to the Linux man page, which states that EINVAL is > returned > if either fd does not refer to a regular file. > Having said that, I do not recall testing the VCHR case under Linux. (ie. > It might > actually work and the man page turns out to be incorrect?) > > I will test this case under Linux when I get home next week, rick > Since there's no standard, I think it's fine for us to support devfs if possible. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365643 - head/bin/cp
Author: asomers
Date: Fri Sep 11 20:49:36 2020
New Revision: 365643
URL: https://svnweb.freebsd.org/changeset/base/365643
Log:
cp: fall back to read/write if copy_file_range fails
Even though copy_file_range has a file-system agnostic version, it still
fails on devfs (perhaps because the file descriptor is non-seekable?) In
that case, fallback to old-fashioned read/write. Fixes
"cp /dev/null /tmp/null"
PR: 249248
Reported by: Michael Butler
Reviewed by: mjg
MFC-With: 365549
Differential Revision:https://reviews.freebsd.org/D26395
Modified:
head/bin/cp/utils.c
Modified: head/bin/cp/utils.c
==
--- head/bin/cp/utils.c Fri Sep 11 20:32:40 2020(r365642)
+++ head/bin/cp/utils.c Fri Sep 11 20:49:36 2020(r365643)
@@ -74,6 +74,26 @@ __FBSDID("$FreeBSD$");
*/
#define BUFSIZE_SMALL (MAXPHYS)
+static int
+copy_fallback(int from_fd, int to_fd, char *buf, size_t bufsize)
+{
+ int rcount;
+ ssize_t wresid, wcount = 0;
+ char *bufp;
+
+ rcount = read(from_fd, buf, bufsize);
+ if (rcount <= 0)
+ return (rcount);
+ for (bufp = buf, wresid = rcount; ; bufp += wcount, wresid -= wcount) {
+ wcount = write(to_fd, bufp, wresid);
+ if (wcount <= 0)
+ break;
+ if (wcount >= (ssize_t)wresid)
+ break;
+ }
+ return (wcount < 0 ? wcount : rcount);
+}
+
int
copy_file(const FTSENT *entp, int dne)
{
@@ -88,6 +108,7 @@ copy_file(const FTSENT *entp, int dne)
#ifdef VM_AND_BUFFER_CACHE_SYNCHRONIZED
char *p;
#endif
+ int use_copy_file_range = 1;
from_fd = to_fd = -1;
if (!lflag && !sflag &&
@@ -212,9 +233,19 @@ copy_file(const FTSENT *entp, int dne)
err(1, "Not enough memory");
}
wtotal = 0;
- while ((rcount = copy_file_range(from_fd, NULL,
- to_fd, NULL, bufsize, 0)) > 0)
- {
+ do {
+ if (use_copy_file_range) {
+ rcount = copy_file_range(from_fd, NULL,
+ to_fd, NULL, bufsize, 0);
+ if (rcount < 0 && errno == EINVAL) {
+ /* Prob a non-seekable FD */
+ use_copy_file_range = 0;
+ }
+ }
+ if (!use_copy_file_range) {
+ rcount = copy_fallback(from_fd, to_fd,
+ buf, bufsize);
+ }
wtotal += rcount;
if (info) {
info = 0;
@@ -223,7 +254,7 @@ copy_file(const FTSENT *entp, int dne)
entp->fts_path, to.p_path,
cp_pct(wtotal, fs->st_size));
}
- }
+ } while (rcount > 0);
if (rcount < 0) {
warn("%s", entp->fts_path);
rval = 1;
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r365614 - in stable/12: sys/dev/virtio/block usr.sbin/bhyve
On Thu, Sep 10, 2020 at 3:01 PM Allan Jude wrote: > Author: allanjude > Date: Thu Sep 10 21:01:22 2020 > New Revision: 365614 > URL: https://svnweb.freebsd.org/changeset/base/365614 > > Log: > MFC r360229, r363255 > > r360229: > Add VIRTIO_BLK_T_DISCARD (TRIM) support to the bhyve virtio-blk backend > > This will advertise support for TRIM to the guest virtio-blk driver and > perform the DIOCGDELETE ioctl on the backing storage if it supports it. > > Thanks to Jason King and others at Joyent and illumos for expanding on > my original patch, adding improvements including better error handling > and making sure to following the virtio spec. > > r363255: > Add VIRTIO_BLK_T_DISCARD support to the virtio-blk driver > > If the hypervisor advertises support for the DISCARD command then the > guest can perform TRIM commands, freeing space on the backing store. > > If VIRTIO_BLK_F_DISCARD is enabled, advertise DISKFLAG_CANDELETE > > Tested with FreeBSD guests on bhyve and KVM > > Relnotes: yes > Sponsored by: Klara Inc. > > Modified: > stable/12/sys/dev/virtio/block/virtio_blk.c > stable/12/sys/dev/virtio/block/virtio_blk.h > stable/12/usr.sbin/bhyve/block_if.c > stable/12/usr.sbin/bhyve/pci_virtio_block.c > Directory Properties: > stable/12/ (props changed) > Yay! ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365549 - head/bin/cp
Author: asomers
Date: Thu Sep 10 02:48:55 2020
New Revision: 365549
URL: https://svnweb.freebsd.org/changeset/base/365549
Log:
cp: use copy_file_range(2)
This has three advantages over write(2)/read(2):
* Fewer context switches and data copies
* Mostly preserves a file's sparseness
* On some file systems (currently NFS 4.2) the file system will perform the
copy in an especially efficient way.
Reviewed by: rmacklem
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D26377
Modified:
head/bin/cp/utils.c
Modified: head/bin/cp/utils.c
==
--- head/bin/cp/utils.c Thu Sep 10 01:49:53 2020(r365548)
+++ head/bin/cp/utils.c Thu Sep 10 02:48:55 2020(r365549)
@@ -212,27 +212,16 @@ copy_file(const FTSENT *entp, int dne)
err(1, "Not enough memory");
}
wtotal = 0;
- while ((rcount = read(from_fd, buf, bufsize)) > 0) {
- for (bufp = buf, wresid = rcount; ;
- bufp += wcount, wresid -= wcount) {
- wcount = write(to_fd, bufp, wresid);
- if (wcount <= 0)
- break;
- wtotal += wcount;
- if (info) {
- info = 0;
- (void)fprintf(stderr,
- "%s -> %s %3d%%\n",
- entp->fts_path, to.p_path,
- cp_pct(wtotal,
fs->st_size));
- }
- if (wcount >= (ssize_t)wresid)
- break;
- }
- if (wcount != (ssize_t)wresid) {
- warn("%s", to.p_path);
- rval = 1;
- break;
+ while ((rcount = copy_file_range(from_fd, NULL,
+ to_fd, NULL, bufsize, 0)) > 0)
+ {
+ wtotal += rcount;
+ if (info) {
+ info = 0;
+ (void)fprintf(stderr,
+ "%s -> %s %3d%%\n",
+ entp->fts_path, to.p_path,
+ cp_pct(wtotal, fs->st_size));
}
}
if (rcount < 0) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365503 - stable/12/usr.bin/nfsstat
Author: asomers
Date: Wed Sep 9 13:16:20 2020
New Revision: 365503
URL: https://svnweb.freebsd.org/changeset/base/365503
Log:
MFC r365262:
Fix output of nfsstat -cE in json or xml mode
Due to a copy/paste error, the "getacl" field was duplicated, but only in
XML or JSON mode, not in txt mode.
Discussed with: rmacklem
Sponsored by: Axcient
Modified:
stable/12/usr.bin/nfsstat/nfsstat.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/usr.bin/nfsstat/nfsstat.c
==
--- stable/12/usr.bin/nfsstat/nfsstat.c Wed Sep 9 12:58:19 2020
(r365502)
+++ stable/12/usr.bin/nfsstat/nfsstat.c Wed Sep 9 13:16:20 2020
(r365503)
@@ -731,7 +731,7 @@ exp_intpr(int clientOnly, int serverOnly, int nfs41)
"{T:PutRootFH/%13.13s}{T:DelegRet/%13.13s}"
"{T:GetAcl/%13.13s}{T:SetAcl/%13.13s}\n");
xo_emit("{:rellckown/%13ju}{:freestateid/%13ju}"
- "{:getacl/%13ju}{:delegret/%13ju}"
+ "{:putrootfh/%13ju}{:delegret/%13ju}"
"{:getacl/%13ju}{:setacl/%13ju}\n",
(uintmax_t)ext_nfsstats.rpccnt[NFSPROC_RELEASELCKOWN],
(uintmax_t)ext_nfsstats.rpccnt[NFSPROC_FREESTATEID],
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365415 - head/share/man/man5
Author: asomers Date: Mon Sep 7 13:44:54 2020 New Revision: 365415 URL: https://svnweb.freebsd.org/changeset/base/365415 Log: nsswitch.conf.5: style fixes Fix some whitespace, and remove the .Tn macro Reported by: mandoc, igor Reviewed by: bcr (manpages) MFC after:1 week Differential Revision:https://reviews.freebsd.org/D26345 Modified: head/share/man/man5/nsswitch.conf.5 Modified: head/share/man/man5/nsswitch.conf.5 == --- head/share/man/man5/nsswitch.conf.5 Mon Sep 7 10:51:48 2020 (r365414) +++ head/share/man/man5/nsswitch.conf.5 Mon Sep 7 13:44:54 2020 (r365415) @@ -1,4 +1,4 @@ -.\"$NetBSD: nsswitch.conf.5,v 1.14 1999/03/17 20:19:47 garbled Exp $ +.\" $NetBSD: nsswitch.conf.5,v 1.14 1999/03/17 20:19:47 garbled Exp $ .\" .\" Copyright (c) 1997, 1998, 1999 The NetBSD Foundation, Inc. .\" All rights reserved. @@ -16,7 +16,7 @@ .\"documentation and/or other materials provided with the distribution. .\" 3. All advertising materials mentioning features or use of this software .\"must display the following acknowledgement: -.\"This product includes software developed by Luke Mewburn. +.\"This product includes software developed by Luke Mewburn. .\" 4. The name of the author may not be used to endorse or promote products .\"derived from this software without specific prior written permission. .\" @@ -48,8 +48,7 @@ file specifies how the .Pp The configuration file controls how a process looks up various databases containing information regarding hosts, users (passwords), groups, etc. -Each database comes from a source (such as local files, DNS, -.Tn NIS , +Each database comes from a source (such as local files, DNS, NIS , and cache), and the order to look up the sources is specified in .Nm . .Pp @@ -182,9 +181,7 @@ Try the next source Return with the current result .El .Ss Format of file -A -.Tn BNF -description of the syntax of +A BNF description of the syntax of .Nm is: .Pp @@ -250,12 +247,10 @@ In historical multi-source implementations, the and .Sq - characters are used to specify the importing of user password and -group information from -.Tn NIS . +group information from NIS . Although .Nm -provides alternative methods of accessing distributed sources such as -.Tn NIS , +provides alternative methods of accessing distributed sources such as NIS , specifying a sole source of .Dq compat will provide the historical behaviour. @@ -328,9 +323,7 @@ resides in To lookup hosts in .Pa /etc/hosts , then in cache, -and then from the DNS, and lookup user information from -.Tn NIS -then files, use: +and then from the DNS, and lookup user information from NIS then files, use: .Pp .Bl -tag -width passwd: -compact .It hosts: @@ -362,9 +355,7 @@ entries. .Fx Ns 's .Lb libc provides stubs for compatibility with NSS modules -written for the -.Tn GNU -C Library +written for the GNU C Library .Nm nsswitch interface. However, these stubs only support the use of the @@ -390,10 +381,8 @@ Project, where it appeared first in .Sh AUTHORS .An Luke Mewburn Aq Mt lu...@netbsd.org wrote this freely distributable name-service switch implementation, -using ideas from the -.Tn ULTRIX +using ideas from the ULTRIX .Xr svc.conf 5 -and -.Tn Solaris +and Solaris .Xr nsswitch.conf 4 manual pages. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364637 - head/sys/kern
On Mon, Aug 24, 2020 at 3:01 AM Mateusz Guzik wrote: > Author: mjg > Date: Mon Aug 24 09:00:57 2020 > New Revision: 364637 > URL: https://svnweb.freebsd.org/changeset/base/364637 > > Log: > cache: lockless reverse lookup > > This enables fully scalable operation for getcwd and significantly > improves > realpath. > > For example: > PATH_CUSTOM=/usr/src ./getcwd_processes -t 104 > before: 1550851 > after: 380135380 > > Tested by:pho > > Modified: > head/sys/kern/vfs_cache.c > > Modified: head/sys/kern/vfs_cache.c > > == > --- head/sys/kern/vfs_cache.c Mon Aug 24 09:00:07 2020(r364636) > +++ head/sys/kern/vfs_cache.c Mon Aug 24 09:00:57 2020(r364637) > @@ -477,6 +485,8 @@ STATNODE_COUNTER(shrinking_skipped, > static void cache_zap_locked(struct namecache *ncp); > static int vn_fullpath_hardlink(struct nameidata *ndp, char **retbuf, > char **freebuf, size_t *buflen); > +static int vn_fullpath_any_smr(struct vnode *vp, struct vnode *rdir, char > *buf, > +char **retbuf, size_t *buflen, bool slash_prefixed, size_t addend); > static int vn_fullpath_any(struct vnode *vp, struct vnode *rdir, char > *buf, > char **retbuf, size_t *buflen); > static int vn_fullpath_dir(struct vnode *vp, struct vnode *rdir, char > *buf, > @@ -2476,9 +2486,17 @@ vn_getcwd(char *buf, char **retbuf, size_t *buflen) > What does the "smr" stand for? -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365391 - head/share/man/man5
Author: asomers Date: Sun Sep 6 20:32:13 2020 New Revision: 365391 URL: https://svnweb.freebsd.org/changeset/base/365391 Log: nsswitch.conf(5): recommend placing cache after files When cache precedes files, and nscd is configured to allow negative caching, commands like "pw groupadd" can fail. The sequence of events looks like: 1. A command like pkg(8) looks up the group, and finds it absent. 2. pkg invokes pw(8) to add the group 3. pkg queries the group, but nscd says it doesn't exist, since it has a negative cache entry for that group. See also: https://lists.freebsd.org/pipermail/freebsd-current/2012-January/031595.html Reviewed by: bcr (manpages) MFC after:1 week Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D26184 Modified: head/share/man/man5/nsswitch.conf.5 Modified: head/share/man/man5/nsswitch.conf.5 == --- head/share/man/man5/nsswitch.conf.5 Sun Sep 6 20:03:13 2020 (r365390) +++ head/share/man/man5/nsswitch.conf.5 Sun Sep 6 20:32:13 2020 (r365391) @@ -33,7 +33,7 @@ .\" .\" $FreeBSD$ .\" -.Dd November 10, 2018 +.Dd September 6, 2020 .Dt NSSWITCH.CONF 5 .Os .Sh NAME @@ -224,20 +224,24 @@ and continue on anything else (i.e, .Ss Cache You can enable caching for the particular database by specifying .Dq cache -as the first source in the +in the .Nm file. +It should come after +.Dq files , +but before remote sources like +.Dq nis . You should also enable caching for this database in .Xr nscd.conf 5 . -If for the particular query +If for a particular query .Dq cache -source returns success, no further sources are queried. +source returns success, then no further sources are queried. On the other hand, if there are no previously cached data, the query result will be placed into the cache right after all other sources are processed. -Note, that +Note that .Dq cache -requires +requires the .Xr nscd 8 daemon to be running. .Ss Compat mode: +/- syntax @@ -321,15 +325,16 @@ resides in .Pa /etc . .El .Sh EXAMPLES -To lookup hosts in cache, then in +To lookup hosts in .Pa /etc/hosts +, then in cache, and then from the DNS, and lookup user information from .Tn NIS then files, use: .Pp .Bl -tag -width passwd: -compact .It hosts: -cache files dns +files cache dns .It passwd: nis [notfound=return] files .It group: ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365389 - head/sys/crypto/via
Author: asomers
Date: Sun Sep 6 19:25:31 2020
New Revision: 365389
URL: https://svnweb.freebsd.org/changeset/base/365389
Log:
padlock(4): fix instapanics with geli authentication
cryptodev_process implementations are supposed to return 0
PR: 247986
Submitted by: jhb
MFC after:1 week
Modified:
head/sys/crypto/via/padlock.c
Modified: head/sys/crypto/via/padlock.c
==
--- head/sys/crypto/via/padlock.c Sun Sep 6 19:03:19 2020
(r365388)
+++ head/sys/crypto/via/padlock.c Sun Sep 6 19:25:31 2020
(r365389)
@@ -275,7 +275,7 @@ out:
#endif
crp->crp_etype = error;
crypto_done(crp);
- return (error);
+ return (0);
}
static device_method_t padlock_methods[] = {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r365262 - head/usr.bin/nfsstat
Author: asomers
Date: Wed Sep 2 17:36:30 2020
New Revision: 365262
URL: https://svnweb.freebsd.org/changeset/base/365262
Log:
Fix output of nfsstat -cE in json or xml mode
Due to a copy/paste error, the "getacl" field was duplicated, but only in
XML or JSON mode, not in txt mode.
Discussed with: rmacklem
MFC after:1 week
Sponsored by: Axcient
Modified:
head/usr.bin/nfsstat/nfsstat.c
Modified: head/usr.bin/nfsstat/nfsstat.c
==
--- head/usr.bin/nfsstat/nfsstat.c Wed Sep 2 17:31:06 2020
(r365261)
+++ head/usr.bin/nfsstat/nfsstat.c Wed Sep 2 17:36:30 2020
(r365262)
@@ -721,7 +721,7 @@ exp_intpr(int clientOnly, int serverOnly, int nfs41)
"{T:PutRootFH/%13.13s}{T:DelegRet/%13.13s}"
"{T:GetAcl/%13.13s}{T:SetAcl/%13.13s}\n");
xo_emit("{:rellckown/%13ju}{:freestateid/%13ju}"
- "{:getacl/%13ju}{:delegret/%13ju}"
+ "{:putrootfh/%13ju}{:delegret/%13ju}"
"{:getacl/%13ju}{:setacl/%13ju}\n",
(uintmax_t)ext_nfsstats.rpccnt[NFSPROC_RELEASELCKOWN],
(uintmax_t)ext_nfsstats.rpccnt[NFSPROC_FREESTATEID],
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r364978 - stable/11/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Sun Aug 30 18:21:54 2020
New Revision: 364978
URL: https://svnweb.freebsd.org/changeset/base/364978
Log:
MFC r364412:
zfs: fix EIO accessing dataset after resuming interrupted receive
ZFS unmounts a dataset while receiving into it and remounts it afterwards.
But if ZFS is resuming an incomplete receive, it screws up and ends up with
a dataset that is mounted, but returns EIO for every access. This commit
fixes that condition.
While the vulnerable code also exists in OpenZFS, the problem is not
reproducible there. Apparently OpenZFS doesn't unmount the destination
dataset during receive, like FreeBSD does.
PR: 248606
Reviewed by: mmacy
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D26034
Modified:
stable/11/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Directory Properties:
stable/11/ (props changed)
Modified: stable/11/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- stable/11/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sun Aug 30 17:40:59 2020(r364977)
+++ stable/11/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sun Aug 30 18:21:54 2020(r364978)
@@ -3415,7 +3415,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
- stream_wantsnewfs) {
+ (stream_wantsnewfs || resuming)) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, 0);
if (clp == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r364974 - stable/12/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Sun Aug 30 16:27:58 2020
New Revision: 364974
URL: https://svnweb.freebsd.org/changeset/base/364974
Log:
MFC r364412:
zfs: fix EIO accessing dataset after resuming interrupted receive
ZFS unmounts a dataset while receiving into it and remounts it afterwards.
But if ZFS is resuming an incomplete receive, it screws up and ends up with
a dataset that is mounted, but returns EIO for every access. This commit
fixes that condition.
While the vulnerable code also exists in OpenZFS, the problem is not
reproducible there. Apparently OpenZFS doesn't unmount the destination
dataset during receive, like FreeBSD does.
PR: 248606
Reviewed by: mmacy
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D26034
Modified:
stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sun Aug 30 07:34:32 2020(r364973)
+++ stable/12/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Sun Aug 30 16:27:58 2020(r364974)
@@ -3433,7 +3433,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
- stream_wantsnewfs) {
+ (stream_wantsnewfs || resuming)) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, 0);
if (clp == NULL) {
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364833 - head/sys/dev/sec
On Wed, Aug 26, 2020 at 1:30 PM Brandon Bergren wrote: > Author: bdragon > Date: Wed Aug 26 19:30:42 2020 > New Revision: 364833 > URL: https://svnweb.freebsd.org/changeset/base/364833 > > Log: > [PowerPC] Fix build failure in sec.c > > Fix a typo in r364799 that was breaking powerpc and powerpcspe build. > > MFC with: 364799 > > Modified: > head/sys/dev/sec/sec.c > Thanks Brandon. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364799 - in head: share/man/man9 sys/crypto/ccp sys/dev/cxgbe/crypto sys/dev/sec sys/kern sys/opencrypto
On Wed, Aug 26, 2020 at 1:30 PM Ed Maste wrote: > On Wed, 26 Aug 2020 at 15:27, Alan Somers wrote: > > > > It probably came copy/pasted from another file :( . I'll fix it; and > figure out why my universe build didn't catch this. > > IIRC powerpcspe is not included in universe/tinderbox. > Oh, it's only for powerpcspe and not for all powerpc? Now I don't feel so bad. -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364799 - in head: share/man/man9 sys/crypto/ccp sys/dev/cxgbe/crypto sys/dev/sec sys/kern sys/opencrypto
On Wed, Aug 26, 2020 at 1:23 PM Brandon Bergren wrote: > > > No, I mean literally. What scope is cb coming from? It's not a variable > > that is in scope for that function as far as I can tell. > > > > Naturally, about two seconds after I sent this, I realized that you meant > that it was a powerpc-specific driver and that's why the other platforms > don't catch it. > > Will fix it myself. > > -- > Brandon Bergren > bdra...@freebsd.org > It probably came copy/pasted from another file :( . I'll fix it; and figure out why my universe build didn't catch this. -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r364800 - head/sys/geom/eli
Author: asomers
Date: Wed Aug 26 02:44:35 2020
New Revision: 364800
URL: https://svnweb.freebsd.org/changeset/base/364800
Log:
geli: use unmapped I/O
Use unmapped I/O for geli. Unlike most geom providers, geli needs to
manipulate data on every read or write. Previously it would always map bios.
On my 16-core, dual socket server using geli atop md(4) devices, with 512B
sectors, this change increases geli IOPs by about 3x.
Note that geli still can't use unmapped I/O when data integrity verification
is enabled (but it could, with a little more work). And it can't use
unmapped I/O in combination with ZFS, because ZFS uses mapped bios.
Reviewed by: markj, kib, jhb, mjg, mat, bcr (manpages)
MFC after:1 week
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25671
Modified:
head/sys/geom/eli/g_eli.c
head/sys/geom/eli/g_eli_privacy.c
Modified: head/sys/geom/eli/g_eli.c
==
--- head/sys/geom/eli/g_eli.c Wed Aug 26 02:37:42 2020(r364799)
+++ head/sys/geom/eli/g_eli.c Wed Aug 26 02:44:35 2020(r364800)
@@ -49,6 +49,8 @@ __FBSDID("$FreeBSD$");
#include
#include
+#include
+
#include
#include
@@ -972,6 +974,15 @@ g_eli_create(struct gctl_req *req, struct g_class *mp,
*/
pp = g_new_providerf(gp, "%s%s", bpp->name, G_ELI_SUFFIX);
pp->flags |= G_PF_DIRECT_SEND | G_PF_DIRECT_RECEIVE;
+ if (CRYPTO_HAS_VMPAGE) {
+ /*
+* On DMAP architectures we can use unmapped I/O. But don't
+* use it with data integrity verification. That code hasn't
+* been written yet.
+*/
+if ((sc->sc_flags & G_ELI_FLAG_AUTH) == 0)
+ pp->flags |= G_PF_ACCEPT_UNMAPPED;
+ }
pp->mediasize = sc->sc_mediasize;
pp->sectorsize = sc->sc_sectorsize;
LIST_FOREACH(gap, >aliases, ga_next)
Modified: head/sys/geom/eli/g_eli_privacy.c
==
--- head/sys/geom/eli/g_eli_privacy.c Wed Aug 26 02:37:42 2020
(r364799)
+++ head/sys/geom/eli/g_eli_privacy.c Wed Aug 26 02:44:35 2020
(r364800)
@@ -63,6 +63,28 @@ __FBSDID("$FreeBSD$");
MALLOC_DECLARE(M_ELI);
/*
+ * Copy data from a (potentially unmapped) bio to a kernelspace buffer.
+ *
+ * The buffer must have at least as much room as bp->bio_length.
+ */
+static void
+g_eli_bio_copyin(struct bio *bp, void *kaddr)
+{
+ struct uio uio;
+ struct iovec iov[1];
+
+ iov[0].iov_base = kaddr;
+ iov[0].iov_len = bp->bio_length;
+ uio.uio_iov = iov;
+ uio.uio_iovcnt = 1;
+ uio.uio_offset = 0;
+ uio.uio_resid = bp->bio_length;
+ uio.uio_segflg = UIO_SYSSPACE;
+ uio.uio_rw = UIO_READ;
+ uiomove_fromphys(bp->bio_ma, bp->bio_ma_offset, bp->bio_length, );
+}
+
+/*
* The function is called after we read and decrypt data.
*
* g_eli_start -> g_eli_crypto_read -> g_io_request -> g_eli_read_done ->
g_eli_crypto_run -> G_ELI_CRYPTO_READ_DONE -> g_io_deliver
@@ -98,8 +120,7 @@ g_eli_crypto_read_done(struct cryptop *crp)
*/
if (bp->bio_inbed < bp->bio_children)
return (0);
- free(bp->bio_driver2, M_ELI);
- bp->bio_driver2 = NULL;
+
if (bp->bio_error != 0) {
G_ELI_LOGREQ(0, bp, "Crypto READ request failed (error=%d).",
bp->bio_error);
@@ -167,6 +188,11 @@ g_eli_crypto_write_done(struct cryptop *crp)
return (0);
}
cbp->bio_data = bp->bio_driver2;
+ /*
+* Clear BIO_UNMAPPED, which was inherited from where we cloned the bio
+* in g_eli_start, because we manually set bio_data
+*/
+ cbp->bio_flags &= ~BIO_UNMAPPED;
cbp->bio_done = g_eli_write_done;
cp = LIST_FIRST(>consumer);
cbp->bio_to = cp->provider;
@@ -236,10 +262,12 @@ g_eli_crypto_run(struct g_eli_worker *wr, struct bio *
{
struct g_eli_softc *sc;
struct cryptop *crp;
+ vm_page_t *pages;
u_int i, nsec, secsize;
off_t dstoff;
- u_char *data;
+ u_char *data = NULL;
int error;
+ int pages_offset;
G_ELI_LOGREQ(3, bp, "%s", __func__);
@@ -258,16 +286,37 @@ g_eli_crypto_run(struct g_eli_worker *wr, struct bio *
if (bp->bio_cmd == BIO_WRITE) {
data = malloc(bp->bio_length, M_ELI, M_WAITOK);
bp->bio_driver2 = data;
- bcopy(bp->bio_data, data, bp->bio_length);
- } else
- data = bp->bio_data;
+ /*
+* This copy could be eliminated by using crypto's output
+* buffer, instead of using a single overwriting buffer.
+*/
+ if ((bp->bio_flags & BIO_UNMAPPED) != 0)
+
svn commit: r364799 - in head: share/man/man9 sys/crypto/ccp sys/dev/cxgbe/crypto sys/dev/sec sys/kern sys/opencrypto
Author: asomers Date: Wed Aug 26 02:37:42 2020 New Revision: 364799 URL: https://svnweb.freebsd.org/changeset/base/364799 Log: crypto(9): add CRYPTO_BUF_VMPAGE crypto(9) functions can now be used on buffers composed of an array of vm_page_t structures, such as those stored in an unmapped struct bio. It requires the running to kernel to support the direct memory map, so not all architectures can use it. Reviewed by: markj, kib, jhb, mjg, mat, bcr (manpages) MFC after:1 week Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D25671 Modified: head/share/man/man9/crypto_buffer.9 head/share/man/man9/crypto_request.9 head/sys/crypto/ccp/ccp.c head/sys/dev/cxgbe/crypto/t4_crypto.c head/sys/dev/sec/sec.c head/sys/kern/subr_bus_dma.c head/sys/opencrypto/criov.c head/sys/opencrypto/crypto.c head/sys/opencrypto/cryptodev.h head/sys/opencrypto/cryptosoft.c Modified: head/share/man/man9/crypto_buffer.9 == --- head/share/man/man9/crypto_buffer.9 Wed Aug 26 02:13:27 2020 (r364798) +++ head/share/man/man9/crypto_buffer.9 Wed Aug 26 02:37:42 2020 (r364799) @@ -30,7 +30,7 @@ .\" .\" $FreeBSD$ .\" -.Dd May 25, 2020 +.Dd August 12, 2020 .Dt CRYPTO_BUFFER 9 .Os .Sh NAME @@ -197,10 +197,17 @@ A scatter/gather list of kernel buffers as described i .It Dv CRYPTO_BUF_MBUF A network memory buffer as described in .Xr mbuf 9 . +.It Dv CRYPTO_BUF_VMPAGE +A scatter/gather list of +.Vt vm_page_t +structures describing pages in the kernel's address space. +This buffer type is only available if +.Dv CRYPTO_HAS_VMPAGE +is true. .El .Pp The structure also contains the following type-specific fields: -.Bl -tag -width " cb_buf_len" +.Bl -tag -width " cb_vm_page_offset" .It Fa cb_buf A pointer to the start of a .Dv CRYPTO_BUF_CONTIG @@ -219,6 +226,19 @@ A pointer to a .Vt struct uio for .Dv CRYPTO_BUF_UIO . +.It Fa cb_vm_page +A pointer to an array of +.Vt struct vm_page +for +.Dv CRYPTO_BUF_VMPAGE . +.It Fa cb_vm_page_len +The total amount of data included in the +.Fa cb_vm_page +array, in bytes. +.It Fa cb_vm_page_offset +Offset in bytes in the first page of +.Fa cb_vm_page +where valid data begins. .El .Ss Cursors Cursors provide a mechanism for iterating over a data buffer. Modified: head/share/man/man9/crypto_request.9 == --- head/share/man/man9/crypto_request.9Wed Aug 26 02:13:27 2020 (r364798) +++ head/share/man/man9/crypto_request.9Wed Aug 26 02:37:42 2020 (r364799) @@ -30,7 +30,7 @@ .\" .\" $FreeBSD$ .\" -.Dd July 16, 2020 +.Dd August 12, 2020 .Dt CRYPTO_REQUEST 9 .Os .Sh NAME @@ -55,11 +55,15 @@ .Ft void .Fn crypto_use_uio "struct cryptop *crp" "struct uio *uio" .Ft void +.Fn crypto_use_vmpage "struct cryptop *crp" "vm_page_t *pages" "int len" "int offset" +.Ft void .Fn crypto_use_output_buf "struct cryptop *crp" "void *buf" "int len" .Ft void .Fn crypto_use_output_mbuf "struct cryptop *crp" "struct mbuf *m" .Ft void .Fn crypto_use_output_uio "struct cryptop *crp" "struct uio *uio" +.Ft void +.Fn crypto_use_output_vmpage "struct cryptop *crp" "vm_page_t *pages" "int len" "int offset" .Sh DESCRIPTION Each symmetric cryptographic operation in the kernel is described by an instance of @@ -141,7 +145,7 @@ mode requests. All requests must have a valid .Fa crp_buf initialized by one of the following functions: -.Bl -tag -width "Fn crypto_use_mbuf" +.Bl -tag -width "Fn crypto_use_vmpage" .It Fn crypto_use_buf Uses an array of .Fa len @@ -156,12 +160,16 @@ as the data buffer. Uses the scatter/gather list .Fa uio as the data buffer. +.It Fn crypto_use_vmpage +Uses the array of +.Vt vm_page_t +structures as the data buffer. .El .Pp One of the following functions should be used to initialize .Fa crp_obuf for requests that use separate input and output buffers: -.Bl -tag -width "Fn crypto_use_output_mbuf" +.Bl -tag -width "Fn crypto_use_output_vmpage" .It Fn crypto_use_output_buf Uses an array of .Fa len @@ -176,6 +184,10 @@ as the output buffer. Uses the scatter/gather list .Fa uio as the output buffer. +.It Fn crypto_use_output_vmpage +Uses the array of +.Vt vm_page_t +structures as the output buffer. .El .Ss Request Regions Each request describes one or more regions in the data buffers. Modified: head/sys/crypto/ccp/ccp.c == --- head/sys/crypto/ccp/ccp.c Wed Aug 26 02:13:27 2020(r364798) +++ head/sys/crypto/ccp/ccp.c Wed Aug 26 02:37:42 2020(r364799) @@ -107,6 +107,10 @@ ccp_populate_sglist(struct sglist *sg, struct crypto_b case CRYPTO_BUF_CONTIG: error = sglist_append(sg, cb->cb_buf, cb->cb_buf_len); break; + case CRYPTO_BUF_VMPAGE: + error =
svn commit: r364412 - head/cddl/contrib/opensolaris/lib/libzfs/common
Author: asomers
Date: Thu Aug 20 01:31:21 2020
New Revision: 364412
URL: https://svnweb.freebsd.org/changeset/base/364412
Log:
zfs: fix EIO accessing dataset after resuming interrupted receive
ZFS unmounts a dataset while receiving into it and remounts it afterwards.
But if ZFS is resuming an incomplete receive, it screws up and ends up with
a dataset that is mounted, but returns EIO for every access. This commit
fixes that condition.
While the vulnerable code also exists in OpenZFS, the problem is not
reproducible there. Apparently OpenZFS doesn't unmount the destination
dataset during receive, like FreeBSD does.
PR: 248606
Reviewed by: mmacy
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D26034
Modified:
head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
Modified: head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
==
--- head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c Thu Aug
20 00:52:53 2020(r364411)
+++ head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c Thu Aug
20 01:31:21 2020(r364412)
@@ -3434,7 +3434,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
- stream_wantsnewfs) {
+ (stream_wantsnewfs || resuming)) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0,
flags->forceunmount ? MS_FORCE : 0);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r364094 - head/tests/sys/fs/fusefs
Author: asomers
Date: Tue Aug 11 01:09:06 2020
New Revision: 364094
URL: https://svnweb.freebsd.org/changeset/base/364094
Log:
fusefs: fix the FUSE_FORGET unit test after r364064
Thanks to r364064, the name cache now returns a hit where previously it
would miss. Adjust the expectations accordingly.
PR: 248583
Reported by: lwhsu
MFC with: r364064
Modified:
head/tests/sys/fs/fusefs/forget.cc
Modified: head/tests/sys/fs/fusefs/forget.cc
==
--- head/tests/sys/fs/fusefs/forget.cc Tue Aug 11 00:41:48 2020
(r364093)
+++ head/tests/sys/fs/fusefs/forget.cc Tue Aug 11 01:09:06 2020
(r364094)
@@ -116,6 +116,7 @@ TEST_F(Forget, invalidate_names)
int err;
EXPECT_LOOKUP(FUSE_ROOT_ID, DNAME)
+ .Times(2)
.WillRepeatedly(Invoke(
ReturnImmediate([=](auto in __unused, auto& out) {
SET_OUT_HEADER_LEN(out, entry);
@@ -142,7 +143,7 @@ TEST_F(Forget, invalidate_names)
out.body.entry.attr_valid = UINT64_MAX;
out.body.entry.entry_valid = UINT64_MAX;
})));
- expect_forget(dir_ino, 2);
+ expect_forget(dir_ino, 1);
/* Access the file to cache its name */
ASSERT_EQ(0, access(FULLFPATH, F_OK)) << strerror(errno);
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364044 - in head: share/man/man9 sys/compat/linuxkpi/common/src sys/kern sys/security/audit sys/sys
So it's just the struct copy, then? I'm impressed that it made that much difference to remove it. -Alan On Fri, Aug 7, 2020 at 7:40 PM Mateusz Guzik wrote: > or to put differently, the difference is disapperance of vn_stat from > the codepath. > > On 8/8/20, Mateusz Guzik wrote: > > tmpfs_getattr fills > 100 bytes worth of data into struct vattr, which > > then has to be converted to struct stat format. The conversion happens > > to copy some of it as it is and branch on other stuff. tmpfs_stat > > elides the entire process. > > > > On 8/8/20, Alan Somers wrote: > >> On Fri, Aug 7, 2020 at 5:06 PM Mateusz Guzik wrote: > >> > >>> Author: mjg > >>> Date: Fri Aug 7 23:06:40 2020 > >>> New Revision: 364044 > >>> URL: https://svnweb.freebsd.org/changeset/base/364044 > >>> > >>> Log: > >>> vfs: add VOP_STAT > >>> > >>> The current scheme of calling VOP_GETATTR adds avoidable overhead. > >>> > >>> An example with tmpfs doing fstat (ops/s): > >>> before: 7488958 > >>> after: 7913833 > >>> > >>> Reviewed by: kib (previous version) > >>> Differential Revision:https://reviews.freebsd.org/D25910 > >>> > >>> Modified: > >>> head/share/man/man9/Makefile > >>> head/share/man/man9/VOP_ATTRIB.9 > >>> head/sys/compat/linuxkpi/common/src/linux_compat.c > >>> head/sys/kern/vfs_default.c > >>> head/sys/kern/vfs_syscalls.c > >>> head/sys/kern/vfs_vnops.c > >>> head/sys/kern/vnode_if.src > >>> head/sys/security/audit/audit_arg.c > >>> head/sys/sys/vnode.h > >>> > >> > >> What avoidable overhead? The tmpfs_stat handler that you added in your > >> next > >> commit looks pretty much the same as vop_getattr. I'm missing where the > >> performance improvement comes from. Could you please fill me in? > >> -Alan > >> > > > > > > -- > > Mateusz Guzik > > > > > -- > Mateusz Guzik > ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r364044 - in head: share/man/man9 sys/compat/linuxkpi/common/src sys/kern sys/security/audit sys/sys
On Fri, Aug 7, 2020 at 5:06 PM Mateusz Guzik wrote: > Author: mjg > Date: Fri Aug 7 23:06:40 2020 > New Revision: 364044 > URL: https://svnweb.freebsd.org/changeset/base/364044 > > Log: > vfs: add VOP_STAT > > The current scheme of calling VOP_GETATTR adds avoidable overhead. > > An example with tmpfs doing fstat (ops/s): > before: 7488958 > after: 7913833 > > Reviewed by: kib (previous version) > Differential Revision:https://reviews.freebsd.org/D25910 > > Modified: > head/share/man/man9/Makefile > head/share/man/man9/VOP_ATTRIB.9 > head/sys/compat/linuxkpi/common/src/linux_compat.c > head/sys/kern/vfs_default.c > head/sys/kern/vfs_syscalls.c > head/sys/kern/vfs_vnops.c > head/sys/kern/vnode_if.src > head/sys/security/audit/audit_arg.c > head/sys/sys/vnode.h > What avoidable overhead? The tmpfs_stat handler that you added in your next commit looks pretty much the same as vop_getattr. I'm missing where the performance improvement comes from. Could you please fill me in? -Alan ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363622 - head/sys/sys
Author: asomers Date: Mon Jul 27 18:57:28 2020 New Revision: 363622 URL: https://svnweb.freebsd.org/changeset/base/363622 Log: Restrict definition of CTL_P1003_1B_MAXID to the kernel This constant is only used to size an array within the kernel. There are probably no legitimate uses in userland. Worse, since the kernel's array could theoretically change size over time, any use of that symbol in userland wouldn't be forwards compatible to new kernel versions. Reviewed by: jhb MFC after:Never Differential Revision:https://reviews.freebsd.org/D25816 Modified: head/sys/sys/sysctl.h Modified: head/sys/sys/sysctl.h == --- head/sys/sys/sysctl.h Mon Jul 27 18:46:20 2020(r363621) +++ head/sys/sys/sysctl.h Mon Jul 27 18:57:28 2020(r363622) @@ -1096,9 +1096,9 @@ TAILQ_HEAD(sysctl_ctx_list, sysctl_ctx_entry); #defineCTL_P1003_1B_SIGQUEUE_MAX 24 /* int */ #defineCTL_P1003_1B_TIMER_MAX 25 /* int */ -#defineCTL_P1003_1B_MAXID 26 - #ifdef _KERNEL + +#defineCTL_P1003_1B_MAXID 26 /* * Declare some common oids. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363486 - in stable/12: sys/geom/eli tests/sys/geom/class/eli
Author: asomers
Date: Fri Jul 24 18:19:25 2020
New Revision: 363486
URL: https://svnweb.freebsd.org/changeset/base/363486
Log:
MFC r363014:
geli: enable direct dispatch
geli does all of its crypto operations in a separate thread pool, so
g_eli_start, g_eli_read_done, and g_eli_write_done don't actually do very
much work. Enabling direct dispatch eliminates the g_up/g_down bottlenecks,
doubling IOPs on my system. This change does not affect the thread pool.
Reviewed by: markj
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25587
Added:
stable/12/tests/sys/geom/class/eli/reentrancy_test.sh
- copied unchanged from r363014,
head/tests/sys/geom/class/eli/reentrancy_test.sh
Modified:
stable/12/sys/geom/eli/g_eli.c
stable/12/tests/sys/geom/class/eli/Makefile
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/geom/eli/g_eli.c
==
--- stable/12/sys/geom/eli/g_eli.c Fri Jul 24 17:56:17 2020
(r363485)
+++ stable/12/sys/geom/eli/g_eli.c Fri Jul 24 18:19:25 2020
(r363486)
@@ -642,6 +642,7 @@ g_eli_read_metadata(struct g_class *mp, struct g_provi
gp->orphan = g_eli_orphan_spoil_assert;
gp->spoiled = g_eli_orphan_spoil_assert;
cp = g_new_consumer(gp);
+ cp->flags |= G_CF_DIRECT_SEND | G_CF_DIRECT_RECEIVE;
error = g_attach(cp, pp);
if (error != 0)
goto end;
@@ -778,6 +779,7 @@ g_eli_create(struct gctl_req *req, struct g_class *mp,
pp = NULL;
cp = g_new_consumer(gp);
+ cp->flags |= G_CF_DIRECT_SEND | G_CF_DIRECT_RECEIVE;
error = g_attach(cp, bpp);
if (error != 0) {
if (req != NULL) {
@@ -865,6 +867,7 @@ g_eli_create(struct gctl_req *req, struct g_class *mp,
* Create decrypted provider.
*/
pp = g_new_providerf(gp, "%s%s", bpp->name, G_ELI_SUFFIX);
+ pp->flags |= G_PF_DIRECT_SEND | G_PF_DIRECT_RECEIVE;
pp->mediasize = sc->sc_mediasize;
pp->sectorsize = sc->sc_sectorsize;
Modified: stable/12/tests/sys/geom/class/eli/Makefile
==
--- stable/12/tests/sys/geom/class/eli/Makefile Fri Jul 24 17:56:17 2020
(r363485)
+++ stable/12/tests/sys/geom/class/eli/Makefile Fri Jul 24 18:19:25 2020
(r363486)
@@ -16,6 +16,7 @@ ATF_TESTS_SH+=integrity_test
ATF_TESTS_SH+= kill_test
ATF_TESTS_SH+= misc_test
ATF_TESTS_SH+= onetime_test
+ATF_TESTS_SH+= reentrancy_test
ATF_TESTS_SH+= resize_test
ATF_TESTS_SH+= setkey_test
Copied: stable/12/tests/sys/geom/class/eli/reentrancy_test.sh (from r363014,
head/tests/sys/geom/class/eli/reentrancy_test.sh)
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ stable/12/tests/sys/geom/class/eli/reentrancy_test.sh Fri Jul 24
18:19:25 2020(r363486, copy of r363014,
head/tests/sys/geom/class/eli/reentrancy_test.sh)
@@ -0,0 +1,69 @@
+# $FreeBSD$
+
+# Test various operations for geli-on-geli providers, to ensure that geli is
+# reentrant.
+
+. $(atf_get_srcdir)/conf.sh
+
+init_test()
+{
+ cipher=$1
+ aalgo=$2
+ secsize=$3
+ ealgo=${cipher%%:*}
+ keylen=${cipher##*:}
+
+ atf_check dd if=/dev/random of=testdata bs=$secsize count=1 status=none
+ atf_check dd if=/dev/random of=keyfile bs=$secsize count=16 status=none
+
+ # Create the lower geli device
+ atf_check -s exit:0 -e ignore \
+ geli init -B none -a $aalgo -e $ealgo -l $keylen -P -K keyfile \
+ -s $secsize ${md}
+ atf_check geli attach -p -k keyfile ${md}
+ # Create the upper geli device
+ atf_check -s exit:0 -e ignore \
+ geli init -B none -a $aalgo -e $ealgo -l $keylen -P -K keyfile \
+ -s $secsize ${md}.eli
+ atf_check geli attach -p -k keyfile ${md}.eli
+ echo ${md} > layered_md_device
+
+ # Ensure we can read and write.
+ atf_check dd if=testdata of=/dev/${md}.eli.eli bs=$secsize count=1 \
+ status=none
+ atf_check dd if=/dev/${md}.eli.eli of=cmpdata bs=$secsize count=1 \
+ status=none
+ atf_check cmp -s testdata cmpdata
+
+ geli detach ${md}.eli 2>/dev/null
+}
+
+atf_test_case init cleanup
+init_head()
+{
+ atf_set "descr" "Initialize a geli provider on top of another"
+ atf_set "require.user" "root"
+ atf_set "timeout" 600
+}
+init_body()
+{
+ sectors=2
+ geli_test_setup
+
+ for_each_geli_config init_test
+}
+init_cleanup()
+{
+ if [ -f layered_md_device ]; then
+ while read provider; do
+ [ -c /dev/${md}.eli.eli ] && \
+ geli detach $md.eli.eli 2>/dev/null
+
svn commit: r363485 - stable/11/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: asomers Date: Fri Jul 24 17:56:17 2020 New Revision: 363485 URL: https://svnweb.freebsd.org/changeset/base/363485 Log: MFC r362891: Fix page fault in zfsctl_snapdir_getattr Must acquire the z_teardown_lock before accessing the zfsvfs_t object. I can't reproduce this panic on demand, but this looks like the correct solution. PR: 247668 Reviewed by: avg Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D25543 Modified: stable/11/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Directory Properties: stable/11/ (props changed) Modified: stable/11/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c == --- stable/11/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Fri Jul 24 17:45:06 2020(r363484) +++ stable/11/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Fri Jul 24 17:56:17 2020(r363485) @@ -1118,12 +1118,13 @@ zfsctl_snapdir_getattr(ap) vnode_t *vp = ap->a_vp; vattr_t *vap = ap->a_vap; zfsvfs_t *zfsvfs = vp->v_vfsp->vfs_data; - dsl_dataset_t *ds = dmu_objset_ds(zfsvfs->z_os); + dsl_dataset_t *ds; sfs_node_t *node = vp->v_data; uint64_t snap_count; int err; ZFS_ENTER(zfsvfs); + ds = dmu_objset_ds(zfsvfs->z_os); zfsctl_common_getattr(vp, vap); vap->va_ctime = dmu_objset_snap_cmtime(zfsvfs->z_os); vap->va_mtime = vap->va_ctime; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363484 - stable/12/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: asomers Date: Fri Jul 24 17:45:06 2020 New Revision: 363484 URL: https://svnweb.freebsd.org/changeset/base/363484 Log: MFC r362891: Fix page fault in zfsctl_snapdir_getattr Must acquire the z_teardown_lock before accessing the zfsvfs_t object. I can't reproduce this panic on demand, but this looks like the correct solution. PR: 247668 Reviewed by: avg Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D25543 Modified: stable/12/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c == --- stable/12/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Fri Jul 24 17:34:44 2020(r363483) +++ stable/12/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Fri Jul 24 17:45:06 2020(r363484) @@ -1120,12 +1120,13 @@ zfsctl_snapdir_getattr(ap) vnode_t *vp = ap->a_vp; vattr_t *vap = ap->a_vap; zfsvfs_t *zfsvfs = vp->v_vfsp->vfs_data; - dsl_dataset_t *ds = dmu_objset_ds(zfsvfs->z_os); + dsl_dataset_t *ds; sfs_node_t *node = vp->v_data; uint64_t snap_count; int err; ZFS_ENTER(zfsvfs); + ds = dmu_objset_ds(zfsvfs->z_os); zfsctl_common_getattr(vp, vap); vap->va_ctime = dmu_objset_snap_cmtime(zfsvfs->z_os); vap->va_mtime = vap->va_ctime; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363402 - in head: sys/geom/eli tests/sys/geom/class/eli
Author: asomers
Date: Tue Jul 21 19:18:29 2020
New Revision: 363402
URL: https://svnweb.freebsd.org/changeset/base/363402
Log:
Fix geli's null cipher, and add a test case
PR: 247954
Submitted by: jhb (sys), asomers (tests)
Reviewed by: jhb (tests), asomers (sys)
MFC after:2 weeks
Sponsored by: Axcient
Modified:
head/sys/geom/eli/g_eli_integrity.c
head/sys/geom/eli/g_eli_privacy.c
head/tests/sys/geom/class/eli/onetime_test.sh
Modified: head/sys/geom/eli/g_eli_integrity.c
==
--- head/sys/geom/eli/g_eli_integrity.c Tue Jul 21 17:34:05 2020
(r363401)
+++ head/sys/geom/eli/g_eli_integrity.c Tue Jul 21 19:18:29 2020
(r363402)
@@ -536,13 +536,15 @@ g_eli_auth_run(struct g_eli_worker *wr, struct bio *bp
crp->crp_digest_start = 0;
crp->crp_payload_start = sc->sc_alen;
crp->crp_payload_length = data_secsize;
- crp->crp_flags |= CRYPTO_F_IV_SEPARATE;
if ((sc->sc_flags & G_ELI_FLAG_FIRST_KEY) == 0) {
crp->crp_cipher_key = g_eli_key_hold(sc, dstoff,
encr_secsize);
}
- g_eli_crypto_ivgen(sc, dstoff, crp->crp_iv,
- sizeof(crp->crp_iv));
+ if (g_eli_ivlen(sc->sc_ealgo) != 0) {
+ crp->crp_flags |= CRYPTO_F_IV_SEPARATE;
+ g_eli_crypto_ivgen(sc, dstoff, crp->crp_iv,
+ sizeof(crp->crp_iv));
+ }
g_eli_auth_keygen(sc, dstoff, authkey);
crp->crp_auth_key = authkey;
Modified: head/sys/geom/eli/g_eli_privacy.c
==
--- head/sys/geom/eli/g_eli_privacy.c Tue Jul 21 17:34:05 2020
(r363401)
+++ head/sys/geom/eli/g_eli_privacy.c Tue Jul 21 19:18:29 2020
(r363402)
@@ -281,13 +281,15 @@ g_eli_crypto_run(struct g_eli_worker *wr, struct bio *
crp->crp_payload_start = 0;
crp->crp_payload_length = secsize;
- crp->crp_flags |= CRYPTO_F_IV_SEPARATE;
if ((sc->sc_flags & G_ELI_FLAG_SINGLE_KEY) == 0) {
crp->crp_cipher_key = g_eli_key_hold(sc, dstoff,
secsize);
}
- g_eli_crypto_ivgen(sc, dstoff, crp->crp_iv,
- sizeof(crp->crp_iv));
+ if (g_eli_ivlen(sc->sc_ealgo) != 0) {
+ crp->crp_flags |= CRYPTO_F_IV_SEPARATE;
+ g_eli_crypto_ivgen(sc, dstoff, crp->crp_iv,
+ sizeof(crp->crp_iv));
+ }
error = crypto_dispatch(crp);
KASSERT(error == 0, ("crypto_dispatch() failed (error=%d)",
Modified: head/tests/sys/geom/class/eli/onetime_test.sh
==
--- head/tests/sys/geom/class/eli/onetime_test.sh Tue Jul 21 17:34:05
2020(r363401)
+++ head/tests/sys/geom/class/eli/onetime_test.sh Tue Jul 21 19:18:29
2020(r363402)
@@ -130,9 +130,54 @@ onetime_d_cleanup()
geli_test_cleanup
}
+atf_test_case onetime cleanup
+onetime_null_head()
+{
+ atf_set "descr" "geli onetime can use the null cipher"
+ atf_set "require.user" "root"
+}
+onetime_null_body()
+{
+ geli_test_setup
+
+ sectors=100
+
+ dd if=/dev/random of=rnd bs=${MAX_SECSIZE} count=${sectors} status=none
+
+ secsize=512
+ ealgo=${cipher%%:*}
+ keylen=${cipher##*:}
+
+ md=$(attach_md -t malloc -s 100k)
+
+ atf_check -s exit:0 -o ignore -e ignore \
+ geli onetime -e null -s ${secsize} ${md}
+
+ atf_check dd if=rnd of=/dev/${md}.eli bs=${secsize} count=${sectors}
status=none
+
+ md_rnd=`dd if=rnd bs=${secsize} count=${sectors} status=none | md5`
+ atf_check_equal 0 $?
+ md_ddev=`dd if=/dev/${md}.eli bs=${secsize} count=${sectors}
status=none | md5`
+ atf_check_equal 0 $?
+ md_edev=`dd if=/dev/${md} bs=${secsize} count=${sectors} status=none |
md5`
+ atf_check_equal 0 $?
+
+ if [ ${md_rnd} != ${md_ddev} ]; then
+ atf_fail "geli did not return the original data"
+ fi
+ if [ ${md_rnd} != ${md_edev} ]; then
+ atf_fail "geli encrypted the data even with the null cipher"
+ fi
+}
+onetime_null_cleanup()
+{
+ geli_test_cleanup
+}
+
atf_init_test_cases()
{
atf_add_test_case onetime
atf_add_test_case onetime_a
atf_add_test_case onetime_d
+ atf_add_test_case onetime_null
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363398 - head/lib/libc/sys
Author: asomers Date: Tue Jul 21 16:46:40 2020 New Revision: 363398 URL: https://svnweb.freebsd.org/changeset/base/363398 Log: [skip ci] document close_range(2) as async-signal-safe Reviewed by: bcr (manpages) MFC after:2 weeks Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D25513 Modified: head/lib/libc/sys/sigaction.2 Modified: head/lib/libc/sys/sigaction.2 == --- head/lib/libc/sys/sigaction.2 Tue Jul 21 16:21:52 2020 (r363397) +++ head/lib/libc/sys/sigaction.2 Tue Jul 21 16:46:40 2020 (r363398) @@ -28,7 +28,7 @@ .\"From: @(#)sigaction.2 8.2 (Berkeley) 4/3/94 .\" $FreeBSD$ .\" -.Dd June 28, 2018 +.Dd June 29, 2020 .Dt SIGACTION 2 .Os .Sh NAME @@ -569,6 +569,7 @@ Extension Interfaces: .Pp .Fn accept4 , .Fn bindat , +.Fn close_range , .Fn closefrom , .Fn connectat , .Fn eaccess , ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363368 - head/sys/crypto/via
Author: asomers
Date: Mon Jul 20 16:12:14 2020
New Revision: 363368
URL: https://svnweb.freebsd.org/changeset/base/363368
Log:
padlock: fix Via Padlock with 192-bit keys
It's been broken since a typo in r359374
Reviewed by: jhb
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25710
Modified:
head/sys/crypto/via/padlock_cipher.c
Modified: head/sys/crypto/via/padlock_cipher.c
==
--- head/sys/crypto/via/padlock_cipher.cMon Jul 20 14:28:26 2020
(r363367)
+++ head/sys/crypto/via/padlock_cipher.cMon Jul 20 16:12:14 2020
(r363368)
@@ -124,7 +124,7 @@ padlock_cipher_setup(struct padlock_session *ses,
{
union padlock_cw *cw;
- if (csp->csp_cipher_klen != 16 && csp->csp_cipher_klen != 25 &&
+ if (csp->csp_cipher_klen != 16 && csp->csp_cipher_klen != 24 &&
csp->csp_cipher_klen != 32) {
return (EINVAL);
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363361 - head/tests/sys/opencrypto
Author: asomers
Date: Mon Jul 20 12:47:15 2020
New Revision: 363361
URL: https://svnweb.freebsd.org/changeset/base/363361
Log:
tests/sys/opencrypto: use python3
python2 will be EOL soon
Reviewed by: lwhsu, jmg
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25682
Modified:
head/tests/sys/opencrypto/Makefile
head/tests/sys/opencrypto/cryptodev.py
head/tests/sys/opencrypto/cryptotest.py
head/tests/sys/opencrypto/runtests.sh
Modified: head/tests/sys/opencrypto/Makefile
==
--- head/tests/sys/opencrypto/Makefile Mon Jul 20 01:55:19 2020
(r363360)
+++ head/tests/sys/opencrypto/Makefile Mon Jul 20 12:47:15 2020
(r363361)
@@ -14,7 +14,7 @@ ATF_TESTS_C+= blake2_test poly1305_test
TAP_TESTS_SH+= runtests
-TEST_METADATA.runtests+= required_programs="python2"
+TEST_METADATA.runtests+= required_programs="python3"
TEST_METADATA.runtests+= required_user="root"
PYMODULES= cryptodev.py cryptodevh.py cryptotest.py
Modified: head/tests/sys/opencrypto/cryptodev.py
==
--- head/tests/sys/opencrypto/cryptodev.py Mon Jul 20 01:55:19 2020
(r363360)
+++ head/tests/sys/opencrypto/cryptodev.py Mon Jul 20 12:47:15 2020
(r363361)
@@ -1,4 +1,4 @@
-#!/usr/local/bin/python2
+#!/usr/local/bin/python3
#
# Copyright (c) 2014 The FreeBSD Foundation
# Copyright 2014 John-Mark Gurney
@@ -31,7 +31,7 @@
# $FreeBSD$
#
-from __future__ import print_function
+
import array
import binascii
from fcntl import ioctl
Modified: head/tests/sys/opencrypto/cryptotest.py
==
--- head/tests/sys/opencrypto/cryptotest.py Mon Jul 20 01:55:19 2020
(r363360)
+++ head/tests/sys/opencrypto/cryptotest.py Mon Jul 20 12:47:15 2020
(r363361)
@@ -1,4 +1,4 @@
-#!/usr/local/bin/python2
+#!/usr/local/bin/python3
#
# Copyright (c) 2014 The FreeBSD Foundation
# All rights reserved.
@@ -30,7 +30,7 @@
# $FreeBSD$
#
-from __future__ import print_function
+
import binascii
import errno
Modified: head/tests/sys/opencrypto/runtests.sh
==
--- head/tests/sys/opencrypto/runtests.sh Mon Jul 20 01:55:19 2020
(r363360)
+++ head/tests/sys/opencrypto/runtests.sh Mon Jul 20 12:47:15 2020
(r363361)
@@ -30,7 +30,7 @@
# $FreeBSD$
#
-: ${PYTHON=python2}
+: ${PYTHON=python3}
if [ ! -d /usr/local/share/nist-kat ]; then
echo "1..0 # SKIP: nist-kat package not installed for test vectors"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r363255 - head/sys/dev/virtio/block
On Thu, Jul 16, 2020 at 10:32 AM Allan Jude wrote: > Author: allanjude > Date: Thu Jul 16 16:32:16 2020 > New Revision: 363255 > URL: https://svnweb.freebsd.org/changeset/base/363255 > > Log: > Add VIRTIO_BLK_T_DISCARD support to the virtio-blk driver > > If the hypervisor advertises support for the DISCARD command then the > guest can perform TRIM commands, freeing space on the backing store. > > If VIRTIO_BLK_F_DISCARD is enabled, advertise DISKFLAG_CANDELETE > > Tested with FreeBSD guests on bhyve and KVM > +1 ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r363014 - in head: sys/geom/eli tests/sys/geom/class/eli
Author: asomers
Date: Wed Jul 8 17:12:12 2020
New Revision: 363014
URL: https://svnweb.freebsd.org/changeset/base/363014
Log:
geli: enable direct dispatch
geli does all of its crypto operations in a separate thread pool, so
g_eli_start, g_eli_read_done, and g_eli_write_done don't actually do very
much work. Enabling direct dispatch eliminates the g_up/g_down bottlenecks,
doubling IOPs on my system. This change does not affect the thread pool.
Reviewed by: markj
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25587
Added:
head/tests/sys/geom/class/eli/reentrancy_test.sh (contents, props changed)
Modified:
head/sys/geom/eli/g_eli.c
head/tests/sys/geom/class/eli/Makefile
Modified: head/sys/geom/eli/g_eli.c
==
--- head/sys/geom/eli/g_eli.c Wed Jul 8 16:50:47 2020(r363013)
+++ head/sys/geom/eli/g_eli.c Wed Jul 8 17:12:12 2020(r363014)
@@ -734,6 +734,7 @@ g_eli_read_metadata_offset(struct g_class *mp, struct
gp->orphan = g_eli_orphan_spoil_assert;
gp->spoiled = g_eli_orphan_spoil_assert;
cp = g_new_consumer(gp);
+ cp->flags |= G_CF_DIRECT_SEND | G_CF_DIRECT_RECEIVE;
error = g_attach(cp, pp);
if (error != 0)
goto end;
@@ -882,6 +883,7 @@ g_eli_create(struct gctl_req *req, struct g_class *mp,
pp = NULL;
cp = g_new_consumer(gp);
+ cp->flags |= G_CF_DIRECT_SEND | G_CF_DIRECT_RECEIVE;
error = g_attach(cp, bpp);
if (error != 0) {
if (req != NULL) {
@@ -969,6 +971,7 @@ g_eli_create(struct gctl_req *req, struct g_class *mp,
* Create decrypted provider.
*/
pp = g_new_providerf(gp, "%s%s", bpp->name, G_ELI_SUFFIX);
+ pp->flags |= G_PF_DIRECT_SEND | G_PF_DIRECT_RECEIVE;
pp->mediasize = sc->sc_mediasize;
pp->sectorsize = sc->sc_sectorsize;
LIST_FOREACH(gap, >aliases, ga_next)
Modified: head/tests/sys/geom/class/eli/Makefile
==
--- head/tests/sys/geom/class/eli/Makefile Wed Jul 8 16:50:47 2020
(r363013)
+++ head/tests/sys/geom/class/eli/Makefile Wed Jul 8 17:12:12 2020
(r363014)
@@ -17,6 +17,7 @@ ATF_TESTS_SH+=kill_test
ATF_TESTS_SH+= misc_test
ATF_TESTS_SH+= onetime_test
ATF_TESTS_SH+= online_resize_test
+ATF_TESTS_SH+= reentrancy_test
ATF_TESTS_SH+= resize_test
ATF_TESTS_SH+= setkey_test
Added: head/tests/sys/geom/class/eli/reentrancy_test.sh
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/tests/sys/geom/class/eli/reentrancy_test.shWed Jul 8 17:12:12
2020(r363014)
@@ -0,0 +1,69 @@
+# $FreeBSD$
+
+# Test various operations for geli-on-geli providers, to ensure that geli is
+# reentrant.
+
+. $(atf_get_srcdir)/conf.sh
+
+init_test()
+{
+ cipher=$1
+ aalgo=$2
+ secsize=$3
+ ealgo=${cipher%%:*}
+ keylen=${cipher##*:}
+
+ atf_check dd if=/dev/random of=testdata bs=$secsize count=1 status=none
+ atf_check dd if=/dev/random of=keyfile bs=$secsize count=16 status=none
+
+ # Create the lower geli device
+ atf_check -s exit:0 -e ignore \
+ geli init -B none -a $aalgo -e $ealgo -l $keylen -P -K keyfile \
+ -s $secsize ${md}
+ atf_check geli attach -p -k keyfile ${md}
+ # Create the upper geli device
+ atf_check -s exit:0 -e ignore \
+ geli init -B none -a $aalgo -e $ealgo -l $keylen -P -K keyfile \
+ -s $secsize ${md}.eli
+ atf_check geli attach -p -k keyfile ${md}.eli
+ echo ${md} > layered_md_device
+
+ # Ensure we can read and write.
+ atf_check dd if=testdata of=/dev/${md}.eli.eli bs=$secsize count=1 \
+ status=none
+ atf_check dd if=/dev/${md}.eli.eli of=cmpdata bs=$secsize count=1 \
+ status=none
+ atf_check cmp -s testdata cmpdata
+
+ geli detach ${md}.eli 2>/dev/null
+}
+
+atf_test_case init cleanup
+init_head()
+{
+ atf_set "descr" "Initialize a geli provider on top of another"
+ atf_set "require.user" "root"
+ atf_set "timeout" 600
+}
+init_body()
+{
+ sectors=2
+ geli_test_setup
+
+ for_each_geli_config init_test
+}
+init_cleanup()
+{
+ if [ -f layered_md_device ]; then
+ while read provider; do
+ [ -c /dev/${md}.eli.eli ] && \
+ geli detach $md.eli.eli 2>/dev/null
+ done < layered_md_device
+ fi
+ geli_test_cleanup
+}
+
+atf_init_test_cases()
+{
+ atf_add_test_case init
+}
___
svn-src-all@freebsd.org mailing list
svn commit: r362891 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs
Author: asomers Date: Thu Jul 2 13:17:31 2020 New Revision: 362891 URL: https://svnweb.freebsd.org/changeset/base/362891 Log: Fix page fault in zfsctl_snapdir_getattr Must acquire the z_teardown_lock before accessing the zfsvfs_t object. I can't reproduce this panic on demand, but this looks like the correct solution. PR: 247668 Reviewed by: avg MFC after:2 weeks Sponsored by: Axcient Differential Revision:https://reviews.freebsd.org/D25543 Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c == --- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.cThu Jul 2 12:58:07 2020(r362890) +++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.cThu Jul 2 13:17:31 2020(r362891) @@ -1121,12 +1121,13 @@ zfsctl_snapdir_getattr(ap) vnode_t *vp = ap->a_vp; vattr_t *vap = ap->a_vap; zfsvfs_t *zfsvfs = vp->v_vfsp->vfs_data; - dsl_dataset_t *ds = dmu_objset_ds(zfsvfs->z_os); + dsl_dataset_t *ds; sfs_node_t *node = vp->v_data; uint64_t snap_count; int err; ZFS_ENTER(zfsvfs); + ds = dmu_objset_ds(zfsvfs->z_os); zfsctl_common_getattr(vp, vap); vap->va_ctime = dmu_objset_snap_cmtime(zfsvfs->z_os); vap->va_mtime = vap->va_ctime; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r342699 - head/sbin/savecore
On Sun, Jun 28, 2020 at 6:55 PM Mark Johnston wrote:
> On Sun, Jun 28, 2020 at 06:40:59PM -0600, Alan Somers wrote:
> > On Wed, Jan 2, 2019 at 10:09 AM Mark Johnston wrote:
> >
> > > Author: markj
> > > Date: Wed Jan 2 17:09:35 2019
> > > New Revision: 342699
> > > URL: https://svnweb.freebsd.org/changeset/base/342699
> > >
> > > Log:
> > > Capsicumize savecore(8).
> > >
> > > - Use cap_fileargs(3) to open dump devices after entering capability
> > > mode, and use cap_syslog(3) to log messages.
> > > - Use a relative directory fd to open output files.
> > > - Use zdopen(3) to compress kernel dumps in capability mode.
> > >
> > > Reviewed by: cem, oshogbo
> > > MFC after:2 months
> > > Sponsored by: The FreeBSD Foundation
> > > Differential Revision:https://reviews.freebsd.org/D18458
> > >
> > > Modified:
> > > head/sbin/savecore/Makefile
> > > head/sbin/savecore/savecore.c
> > >
> > > Modified: head/sbin/savecore/savecore.c
> > >
> > >
> ==
> > > --- head/sbin/savecore/savecore.c Wed Jan 2 16:42:07 2019
> > > (r342698)
> > > +++ head/sbin/savecore/savecore.c Wed Jan 2 17:09:35 2019
> > > (r342699)
> > >
> > > +static char **
> > > +enum_dumpdevs(int *argcp)
> > > +{
> > > + struct fstab *fsp;
> > > + char **argv;
> > > + int argc, n;
> > > +
> > > + /*
> > > +* We cannot use getfsent(3) in capability mode, so we must
> > > +* scan /etc/fstab and build up a list of candidate devices
> > > +* before proceeding.
> > > +*/
> > > + argc = 0;
> > > + n = 8;
> > > + argv = malloc(n * sizeof(*argv));
> > >
> >
> > It looks like the memory allocated here
> >
> >
> > > + if (argv == NULL) {
> > > + logmsg(LOG_ERR, "malloc(): %m");
> > > + exit(1);
> > > + }
> > > + for (;;) {
> > > + fsp = getfsent();
> > > + if (fsp == NULL)
> > > + break;
> > > + if (strcmp(fsp->fs_vfstype, "swap") != 0 &&
> > > + strcmp(fsp->fs_vfstype, "dump") != 0)
> > > + continue;
> > > + if (argc >= n) {
> > > + n *= 2;
> > > + argv = realloc(argv, n * sizeof(*argv));
> > >
> >
> > and here
> >
> >
> > > + if (argv == NULL) {
> > > + logmsg(LOG_ERR, "realloc(): %m");
> > > + exit(1);
> > > + }
> > > + }
> > > + argv[argc] = strdup(fsp->fs_spec);
> > >
> >
> > and here is leaked. I can't find any corresponding free. However,
> neither
> > Valgrind nor Coverity complains. What am I missing? Does this memory
> > sneakily get freed by a subroutine somewhere, or does Capsicum confuse
> our
> > tools?
>
> I'm not sure why Capsicum would change anything. It would be worth
> testing devel/valgrind-devel with https://reviews.freebsd.org/D25452
> applied, to see if it's able to detect that bug.
>
Using that new valgrind-devel did not find the leaks. However, it _did_
print a helpful error message explaining why Capsicum interferes.
Commenting out caph_enter_casper allowed valgrind to find the leaks. It
also found two more than I didn't know about. Thanks for the tip.
https://www.valgrind.org/docs/manual/manual-core-adv.html#manual-core-adv.clientreq
-Alan
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r362790 - in head: sbin/savecore share/man/man5
Author: asomers
Date: Mon Jun 29 22:12:23 2020
New Revision: 362790
URL: https://svnweb.freebsd.org/changeset/base/362790
Log:
savecore: accept device names without the /dev/ prefix
dumpon has accepted device names without the prefix ever since r291207.
Since dumpon and savecore are always paired, they ought to accept the same
arguments. Prior to this change, specifying 'dumpdev="da3"' in
/etc/rc.conf, for example, would result in dumpon working just fine but
savecore complaining that "Dump device does not exist".
PR: 247618
Reviewed by: cem, bcr
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D25500
Modified:
head/sbin/savecore/savecore.c
head/share/man/man5/rc.conf.5
Modified: head/sbin/savecore/savecore.c
==
--- head/sbin/savecore/savecore.c Mon Jun 29 21:35:50 2020
(r362789)
+++ head/sbin/savecore/savecore.c Mon Jun 29 22:12:23 2020
(r362790)
@@ -973,7 +973,45 @@ closefd:
close(fddev);
}
+/* Prepend "/dev/" to any arguments that don't already have it */
static char **
+devify(int argc, char **argv)
+{
+ char **devs;
+ int i, l;
+
+ devs = malloc(argc * sizeof(*argv));
+ if (devs == NULL) {
+ logmsg(LOG_ERR, "malloc(): %m");
+ exit(1);
+ }
+ for (i = 0; i < argc; i++) {
+ if (strncmp(argv[i], _PATH_DEV, sizeof(_PATH_DEV) - 1) == 0)
+ devs[i] = strdup(argv[i]);
+ else {
+ char *fullpath;
+
+ fullpath = malloc(PATH_MAX);
+ if (fullpath == NULL) {
+ logmsg(LOG_ERR, "malloc(): %m");
+ exit(1);
+ }
+ l = snprintf(fullpath, PATH_MAX, "%s%s", _PATH_DEV,
+ argv[i]);
+ if (l < 0) {
+ logmsg(LOG_ERR, "snprintf(): %m");
+ exit(1);
+ } else if (l >= PATH_MAX) {
+ logmsg(LOG_ERR, "device name too long");
+ exit(1);
+ }
+ devs[i] = fullpath;
+ }
+ }
+ return (devs);
+}
+
+static char **
enum_dumpdevs(int *argcp)
{
struct fstab *fsp;
@@ -1069,6 +1107,7 @@ main(int argc, char **argv)
{
cap_rights_t rights;
const char *savedir;
+ char **devs;
int i, ch, error, savedirfd;
checkfor = compress = clear = force = keep = verbose = 0;
@@ -1132,7 +1171,9 @@ main(int argc, char **argv)
argv++;
}
if (argc == 0)
- argv = enum_dumpdevs();
+ devs = enum_dumpdevs();
+ else
+ devs = devify(argc, argv);
savedirfd = open(savedir, O_RDONLY | O_DIRECTORY);
if (savedirfd < 0) {
@@ -1148,10 +1189,10 @@ main(int argc, char **argv)
}
/* Enter capability mode. */
- init_caps(argc, argv);
+ init_caps(argc, devs);
for (i = 0; i < argc; i++)
- DoFile(savedir, savedirfd, argv[i]);
+ DoFile(savedir, savedirfd, devs[i]);
/* Emit minimal output. */
if (nfound == 0) {
Modified: head/share/man/man5/rc.conf.5
==
--- head/share/man/man5/rc.conf.5 Mon Jun 29 21:35:50 2020
(r362789)
+++ head/share/man/man5/rc.conf.5 Mon Jun 29 22:12:23 2020
(r362790)
@@ -24,7 +24,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd June 23, 2020
+.Dd June 28, 2020
.Dt RC.CONF 5
.Os
.Sh NAME
@@ -3469,7 +3469,9 @@ the first suitable swap device listed in
.Pa /etc/fstab
will be used as dump device.
Otherwise, the value of this variable is passed as the argument to
-.Xr dumpon 8 .
+.Xr dumpon 8
+and
+.Xr savecore 8 .
To disable crash dumps, set this variable to
.Dq Li NO .
.It Va dumpon_flags
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r342699 - head/sbin/savecore
On Sun, Jun 28, 2020 at 6:46 PM Warner Losh wrote:
>
>
> On Sun, Jun 28, 2020, 6:41 PM Alan Somers wrote:
>
>> On Wed, Jan 2, 2019 at 10:09 AM Mark Johnston wrote:
>>
>>> Author: markj
>>> Date: Wed Jan 2 17:09:35 2019
>>> New Revision: 342699
>>> URL: https://svnweb.freebsd.org/changeset/base/342699
>>>
>>> Log:
>>> Capsicumize savecore(8).
>>>
>>> - Use cap_fileargs(3) to open dump devices after entering capability
>>> mode, and use cap_syslog(3) to log messages.
>>> - Use a relative directory fd to open output files.
>>> - Use zdopen(3) to compress kernel dumps in capability mode.
>>>
>>> Reviewed by: cem, oshogbo
>>> MFC after:2 months
>>> Sponsored by: The FreeBSD Foundation
>>> Differential Revision:https://reviews.freebsd.org/D18458
>>>
>>> Modified:
>>> head/sbin/savecore/Makefile
>>> head/sbin/savecore/savecore.c
>>>
>>> Modified: head/sbin/savecore/savecore.c
>>>
>>> ==
>>> --- head/sbin/savecore/savecore.c Wed Jan 2 16:42:07 2019
>>> (r342698)
>>> +++ head/sbin/savecore/savecore.c Wed Jan 2 17:09:35 2019
>>> (r342699)
>>>
>>> +static char **
>>> +enum_dumpdevs(int *argcp)
>>> +{
>>> + struct fstab *fsp;
>>> + char **argv;
>>> + int argc, n;
>>> +
>>> + /*
>>> +* We cannot use getfsent(3) in capability mode, so we must
>>> +* scan /etc/fstab and build up a list of candidate devices
>>> +* before proceeding.
>>> +*/
>>> + argc = 0;
>>> + n = 8;
>>> + argv = malloc(n * sizeof(*argv));
>>>
>>
>> It looks like the memory allocated here
>>
>>
>>> + if (argv == NULL) {
>>> + logmsg(LOG_ERR, "malloc(): %m");
>>> + exit(1);
>>> + }
>>> + for (;;) {
>>> + fsp = getfsent();
>>> + if (fsp == NULL)
>>> + break;
>>> + if (strcmp(fsp->fs_vfstype, "swap") != 0 &&
>>> + strcmp(fsp->fs_vfstype, "dump") != 0)
>>> + continue;
>>> + if (argc >= n) {
>>> + n *= 2;
>>> + argv = realloc(argv, n * sizeof(*argv));
>>>
>>
>> and here
>>
>>
>>> + if (argv == NULL) {
>>> + logmsg(LOG_ERR, "realloc(): %m");
>>> + exit(1);
>>> + }
>>> + }
>>> + argv[argc] = strdup(fsp->fs_spec);
>>>
>>
>> and here is leaked. I can't find any corresponding free. However,
>> neither Valgrind nor Coverity complains. What am I missing? Does this
>> memory sneakily get freed by a subroutine somewhere, or does Capsicum
>> confuse our tools?
>>
>
> So the other spots adjusted large, but this one sets one of its elements.
> Help me understand how that is a leak? I'm sure I'm just confused.
>
Because strdup itself allocates new memory. strdup's return value is
always supposed to be freed.
-Alan
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r342699 - head/sbin/savecore
On Wed, Jan 2, 2019 at 10:09 AM Mark Johnston wrote:
> Author: markj
> Date: Wed Jan 2 17:09:35 2019
> New Revision: 342699
> URL: https://svnweb.freebsd.org/changeset/base/342699
>
> Log:
> Capsicumize savecore(8).
>
> - Use cap_fileargs(3) to open dump devices after entering capability
> mode, and use cap_syslog(3) to log messages.
> - Use a relative directory fd to open output files.
> - Use zdopen(3) to compress kernel dumps in capability mode.
>
> Reviewed by: cem, oshogbo
> MFC after:2 months
> Sponsored by: The FreeBSD Foundation
> Differential Revision:https://reviews.freebsd.org/D18458
>
> Modified:
> head/sbin/savecore/Makefile
> head/sbin/savecore/savecore.c
>
> Modified: head/sbin/savecore/savecore.c
>
> ==
> --- head/sbin/savecore/savecore.c Wed Jan 2 16:42:07 2019
> (r342698)
> +++ head/sbin/savecore/savecore.c Wed Jan 2 17:09:35 2019
> (r342699)
>
> +static char **
> +enum_dumpdevs(int *argcp)
> +{
> + struct fstab *fsp;
> + char **argv;
> + int argc, n;
> +
> + /*
> +* We cannot use getfsent(3) in capability mode, so we must
> +* scan /etc/fstab and build up a list of candidate devices
> +* before proceeding.
> +*/
> + argc = 0;
> + n = 8;
> + argv = malloc(n * sizeof(*argv));
>
It looks like the memory allocated here
> + if (argv == NULL) {
> + logmsg(LOG_ERR, "malloc(): %m");
> + exit(1);
> + }
> + for (;;) {
> + fsp = getfsent();
> + if (fsp == NULL)
> + break;
> + if (strcmp(fsp->fs_vfstype, "swap") != 0 &&
> + strcmp(fsp->fs_vfstype, "dump") != 0)
> + continue;
> + if (argc >= n) {
> + n *= 2;
> + argv = realloc(argv, n * sizeof(*argv));
>
and here
> + if (argv == NULL) {
> + logmsg(LOG_ERR, "realloc(): %m");
> + exit(1);
> + }
> + }
> + argv[argc] = strdup(fsp->fs_spec);
>
and here is leaked. I can't find any corresponding free. However, neither
Valgrind nor Coverity complains. What am I missing? Does this memory
sneakily get freed by a subroutine somewhere, or does Capsicum confuse our
tools?
-Alan
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r362280 - in head/sys/fs: nfs nfsclient nfsserver
Author: asomers
Date: Wed Jun 17 16:20:19 2020
New Revision: 362280
URL: https://svnweb.freebsd.org/changeset/base/362280
Log:
Remove vfs_statfs and vnode_mount macros from NFS
These macro definitions are no longer needed as the NFS OSX port is long
dead. The vfs_statfs macro conflicts with the vfsops field of the same
name.
Submitted by: shivank@
Reviewed by: rmacklem
MFC after:2 weeks
Sponsored by: Google, Inc. (GSoC 2020)
Differential Revision:https://reviews.freebsd.org/D25263
Modified:
head/sys/fs/nfs/nfs_commonsubs.c
head/sys/fs/nfs/nfsdport.h
head/sys/fs/nfs/nfskpiport.h
head/sys/fs/nfs/nfsport.h
head/sys/fs/nfsclient/nfs_clbio.c
head/sys/fs/nfsclient/nfs_clrpcops.c
head/sys/fs/nfsclient/nfs_clstate.c
head/sys/fs/nfsclient/nfs_clvnops.c
head/sys/fs/nfsserver/nfs_nfsdserv.c
Modified: head/sys/fs/nfs/nfs_commonsubs.c
==
--- head/sys/fs/nfs/nfs_commonsubs.cWed Jun 17 15:57:59 2020
(r362279)
+++ head/sys/fs/nfs/nfs_commonsubs.cWed Jun 17 16:20:19 2020
(r362280)
@@ -1402,9 +1402,9 @@ nfsv4_loadattr(struct nfsrv_descript *nd, vnode_t vp,
if (compare) {
if (*retcmpp == 0) {
if (thyp != (u_int64_t)
- vfs_statfs(vnode_mount(vp))->f_fsid.val[0]
||
+ vp->v_mount->mnt_stat.f_fsid.val[0] ||
thyp2 != (u_int64_t)
- vfs_statfs(vnode_mount(vp))->f_fsid.val[1])
+ vp->v_mount->mnt_stat.f_fsid.val[1])
*retcmpp = NFSERR_NOTSAME;
}
} else if (nap != NULL) {
@@ -1876,7 +1876,7 @@ nfsv4_loadattr(struct nfsrv_descript *nd, vnode_t vp,
*/
savuid = p->p_cred->p_ruid;
p->p_cred->p_ruid = cred->cr_uid;
- if (!VFS_QUOTACTL(vnode_mount(vp),QCMD(Q_GETQUOTA,
+ if (!VFS_QUOTACTL(vp->v_mount,QCMD(Q_GETQUOTA,
USRQUOTA), cred->cr_uid, (caddr_t)))
freenum = min(dqb.dqb_bhardlimit, freenum);
p->p_cred->p_ruid = savuid;
@@ -1905,7 +1905,7 @@ nfsv4_loadattr(struct nfsrv_descript *nd, vnode_t vp,
*/
savuid = p->p_cred->p_ruid;
p->p_cred->p_ruid = cred->cr_uid;
- if (!VFS_QUOTACTL(vnode_mount(vp),QCMD(Q_GETQUOTA,
+ if (!VFS_QUOTACTL(vp->v_mount,QCMD(Q_GETQUOTA,
USRQUOTA), cred->cr_uid, (caddr_t)))
freenum = min(dqb.dqb_bsoftlimit, freenum);
p->p_cred->p_ruid = savuid;
@@ -1931,7 +1931,7 @@ nfsv4_loadattr(struct nfsrv_descript *nd, vnode_t vp,
*/
savuid = p->p_cred->p_ruid;
p->p_cred->p_ruid = cred->cr_uid;
- if (!VFS_QUOTACTL(vnode_mount(vp),QCMD(Q_GETQUOTA,
+ if (!VFS_QUOTACTL(vp->v_mount,QCMD(Q_GETQUOTA,
USRQUOTA), cred->cr_uid, (caddr_t)))
freenum = dqb.dqb_curblocks;
p->p_cred->p_ruid = savuid;
Modified: head/sys/fs/nfs/nfsdport.h
==
--- head/sys/fs/nfs/nfsdport.h Wed Jun 17 15:57:59 2020(r362279)
+++ head/sys/fs/nfs/nfsdport.h Wed Jun 17 16:20:19 2020(r362280)
@@ -101,12 +101,6 @@ struct nfsexstuff {
(n)->cn_flags = (f);\
} while (0)
-/*
- * A little bit of Darwin vfs kpi.
- */
-#definevnode_mount(v) ((v)->v_mount)
-#definevfs_statfs(m) (&((m)->mnt_stat))
-
#defineNFSPATHLEN_Tsize_t
/*
Modified: head/sys/fs/nfs/nfskpiport.h
==
--- head/sys/fs/nfs/nfskpiport.hWed Jun 17 15:57:59 2020
(r362279)
+++ head/sys/fs/nfs/nfskpiport.hWed Jun 17 16:20:19 2020
(r362280)
@@ -36,11 +36,9 @@
* Darwin8 and hopefully subsequent releases from Apple.)
*/
typedefstruct mount * mount_t;
-#definevfs_statfs(m) (&((m)->mnt_stat))
#definevfs_flags(m)((m)->mnt_flag)
typedef struct vnode * vnode_t;
-#definevnode_mount(v) ((v)->v_mount)
#definevnode_vtype(v) ((v)->v_type)
#endif /* _NFS_NFSKPIPORT_H */
Modified: head/sys/fs/nfs/nfsport.h
svn commit: r362118 - stable/12/sys/geom/eli
Author: asomers
Date: Fri Jun 12 20:39:42 2020
New Revision: 362118
URL: https://svnweb.freebsd.org/changeset/base/362118
Log:
MFC r361562:
geli: fix a livelock during panic
During any kind of shutdown, kern_reboot calls geli's pre_sync event hook,
which tries to destroy all unused geli devices. But during a panic, geli
can't destroy any devices, because the scheduler is stopped, so it can't
switch threads. A livelock results, and the system never dumps core.
This commit fixes the problem by refusing to destroy any devices during
panic, used or otherwise.
PR: 246207
Reviewed by: jhb
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D24697
Modified:
stable/12/sys/geom/eli/g_eli.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/geom/eli/g_eli.c
==
--- stable/12/sys/geom/eli/g_eli.c Fri Jun 12 20:33:00 2020
(r362117)
+++ stable/12/sys/geom/eli/g_eli.c Fri Jun 12 20:39:42 2020
(r362118)
@@ -1320,11 +1320,13 @@ g_eli_shutdown_pre_sync(void *arg, int howto)
continue;
pp = LIST_FIRST(>provider);
KASSERT(pp != NULL, ("No provider? gp=%p (%s)", gp, gp->name));
- if (pp->acr + pp->acw + pp->ace == 0)
- error = g_eli_destroy(sc, TRUE);
- else {
+ if (pp->acr != 0 || pp->acw != 0 || pp->ace != 0 ||
+ SCHEDULER_STOPPED())
+ {
sc->sc_flags |= G_ELI_FLAG_RW_DETACH;
gp->access = g_eli_access;
+ } else {
+ error = g_eli_destroy(sc, TRUE);
}
}
g_topology_unlock();
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r362117 - stable/12/share/man/man4
Author: asomers Date: Fri Jun 12 20:33:00 2020 New Revision: 362117 URL: https://svnweb.freebsd.org/changeset/base/362117 Log: MFC r361439: [skip ci] ip.4: fix typos Modified: stable/12/share/man/man4/ip.4 Directory Properties: stable/12/ (props changed) Modified: stable/12/share/man/man4/ip.4 == --- stable/12/share/man/man4/ip.4 Fri Jun 12 20:32:26 2020 (r362116) +++ stable/12/share/man/man4/ip.4 Fri Jun 12 20:33:00 2020 (r362117) @@ -28,7 +28,7 @@ .\" @(#)ip.4 8.2 (Berkeley) 11/30/93 .\" $FreeBSD$ .\" -.Dd February 22, 2019 +.Dd May 24, 2020 .Dt IP 4 .Os .Sh NAME @@ -144,7 +144,7 @@ the .Xr recvmsg 2 call will return the destination .Tn IP -address and destination port or a +address and destination port for a .Tn UDP datagram. The @@ -155,13 +155,13 @@ structure points to a buffer that contains a .Vt cmsghdr structure followed by the -.Tn in_sockkaddr -structre. +.Tn sockaddr_in +structure. The .Vt cmsghdr fields have the following values: .Bd -literal -cmsg_len = CMSG_LEN(sizeof(struct in_sockaddr)) +cmsg_len = CMSG_LEN(sizeof(struct sockaddr_in)) cmsg_level = IPPROTO_IP cmsg_type = IP_ORIGDSTADDR .Ed ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r362116 - in stable/12: sys/fs/fuse tests/sys/fs/fusefs
Author: asomers
Date: Fri Jun 12 20:32:26 2020
New Revision: 362116
URL: https://svnweb.freebsd.org/changeset/base/362116
Log:
MFC r361401:
Fix issues with FUSE_ACCESS when default_permissions is disabled
This patch fixes two issues relating to FUSE_ACCESS when the
default_permissions mount option is disabled:
* VOP_ACCESS() calls with VADMIN set should never be sent to a fuse server
in the form of FUSE_ACCESS operations. The FUSE protocol has no equivalent
of VADMIN, so we must evaluate such things kernel-side, regardless of the
default_permissions setting.
* The FUSE protocol only requires FUSE_ACCESS to be sent for two purposes:
for the access(2) syscall and to check directory permissions for
searchability during lookup. FreeBSD sends it much more frequently, due to
differences between our VFS and Linux's, for which FUSE was designed. But
this patch does eliminate several cases not required by the FUSE protocol:
* for any FUSE_*XATTR operation
* when creating a new file
* when deleting a file
* when setting timestamps, such as by utimensat(2).
* Additionally, when default_permissions is disabled, this patch removes one
FUSE_GETATTR operation when deleting a file.
PR: 245689
Reported by: MooseFS FreeBSD Team
Reviewed by: cem
Differential Revision:https://reviews.freebsd.org/D24777
Modified:
stable/12/sys/fs/fuse/fuse_internal.c
stable/12/sys/fs/fuse/fuse_vnops.c
stable/12/tests/sys/fs/fusefs/access.cc
stable/12/tests/sys/fs/fusefs/rename.cc
stable/12/tests/sys/fs/fusefs/rmdir.cc
stable/12/tests/sys/fs/fusefs/unlink.cc
stable/12/tests/sys/fs/fusefs/utils.cc
stable/12/tests/sys/fs/fusefs/utils.hh
stable/12/tests/sys/fs/fusefs/xattr.cc
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/fs/fuse/fuse_internal.c
==
--- stable/12/sys/fs/fuse/fuse_internal.c Fri Jun 12 20:27:37 2020
(r362115)
+++ stable/12/sys/fs/fuse/fuse_internal.c Fri Jun 12 20:32:26 2020
(r362116)
@@ -158,6 +158,7 @@ fuse_internal_get_cached_vnode(struct mount* mp, ino_t
return 0;
}
+SDT_PROBE_DEFINE0(fusefs, , internal, access_vadmin);
/* Synchronously send a FUSE_ACCESS operation */
int
fuse_internal_access(struct vnode *vp,
@@ -210,10 +211,18 @@ fuse_internal_access(struct vnode *vp,
va.va_gid, mode, cred, NULL);
}
+ if (mode & VADMIN) {
+ /*
+* The FUSE protocol doesn't have an equivalent of VADMIN, so
+* it's a bug if we ever reach this point with that bit set.
+*/
+ SDT_PROBE0(fusefs, , internal, access_vadmin);
+ }
+
if (!fsess_isimpl(mp, FUSE_ACCESS))
return 0;
- if ((mode & (VWRITE | VAPPEND | VADMIN)) != 0)
+ if ((mode & (VWRITE | VAPPEND)) != 0)
mask |= W_OK;
if ((mode & VREAD) != 0)
mask |= R_OK;
Modified: stable/12/sys/fs/fuse/fuse_vnops.c
==
--- stable/12/sys/fs/fuse/fuse_vnops.c Fri Jun 12 20:27:37 2020
(r362115)
+++ stable/12/sys/fs/fuse/fuse_vnops.c Fri Jun 12 20:32:26 2020
(r362116)
@@ -233,6 +233,7 @@ fuse_extattr_check_cred(struct vnode *vp, int ns, stru
{
struct mount *mp = vnode_mount(vp);
struct fuse_data *data = fuse_get_mpdata(mp);
+ int default_permissions = data->dataflags & FSESS_DEFAULT_PERMISSIONS;
/*
* Kernel-invoked always succeeds.
@@ -246,13 +247,16 @@ fuse_extattr_check_cred(struct vnode *vp, int ns, stru
*/
switch (ns) {
case EXTATTR_NAMESPACE_SYSTEM:
- if (data->dataflags & FSESS_DEFAULT_PERMISSIONS) {
+ if (default_permissions) {
return (priv_check_cred(cred, PRIV_VFS_EXTATTR_SYSTEM,
0));
}
- /* FALLTHROUGH */
+ return (0);
case EXTATTR_NAMESPACE_USER:
- return (fuse_internal_access(vp, accmode, td, cred));
+ if (default_permissions) {
+ return (fuse_internal_access(vp, accmode, td, cred));
+ }
+ return (0);
default:
return (EPERM);
}
@@ -984,6 +988,8 @@ fuse_vnop_lookup(struct vop_lookup_args *ap)
int wantparent = flags & (LOCKPARENT | WANTPARENT);
int islastcn = flags & ISLASTCN;
struct mount *mp = vnode_mount(dvp);
+ struct fuse_data *data = fuse_get_mpdata(mp);
+ int default_permissions = data->dataflags & FSESS_DEFAULT_PERMISSIONS;
int err = 0;
int lookup_err = 0;
@@ -1107,7 +1113,11 @@ fuse_vnop_lookup(struct vop_lookup_args *ap)
if (lookup_err) {
/* Entry not
svn commit: r362115 - stable/12/sys/fs/fuse
Author: asomers Date: Fri Jun 12 20:27:37 2020 New Revision: 362115 URL: https://svnweb.freebsd.org/changeset/base/362115 Log: MFC r361399: Disable nullfs cacheing on top of fusefs Nullfs cacheing can keep a large number of vnodes active. That results in more active FUSE file handles, causing some FUSE servers to use extra resources. Disable nullfs cacheing for fusefs, just like we already do for NFSv4. PR: 245688 Reported by: MooseFS FreeBSD Team Modified: stable/12/sys/fs/fuse/fuse_vfsops.c Directory Properties: stable/12/ (props changed) Modified: stable/12/sys/fs/fuse/fuse_vfsops.c == --- stable/12/sys/fs/fuse/fuse_vfsops.c Fri Jun 12 20:11:25 2020 (r362114) +++ stable/12/sys/fs/fuse/fuse_vfsops.c Fri Jun 12 20:27:37 2020 (r362115) @@ -425,6 +425,11 @@ fuse_vfsop_mount(struct mount *mp) */ mp->mnt_flag &= ~MNT_LOCAL; mp->mnt_kern_flag |= MNTK_USES_BCACHE; + /* +* Disable nullfs cacheing because it can consume too many resources in +* the FUSE server. +*/ + mp->mnt_kern_flag |= MNTK_NULL_NOCACHE; MNT_IUNLOCK(mp); /* We need this here as this slot is used by getnewvnode() */ mp->mnt_stat.f_iosize = maxbcachebuf; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r362114 - stable/12/tests/sys/fs/fusefs
Author: asomers
Date: Fri Jun 12 20:11:25 2020
New Revision: 362114
URL: https://svnweb.freebsd.org/changeset/base/362114
Log:
MFC r361223:
fusefs: fix intermittency in some ENOENT tests
When a FUSE operation other than LOOKUP returns ENOENT, the kernel will
reclaim that vnode, resuling in a FUSE_FORGET being sent a short while
later. Many of the ENOENT tests weren't expecting those FUSE_FORGET
operations. They usually passed by luck since FUSE_FORGET is often delayed.
This commit adds appropriate expectations.
Modified:
stable/12/tests/sys/fs/fusefs/getattr.cc
stable/12/tests/sys/fs/fusefs/open.cc
stable/12/tests/sys/fs/fusefs/opendir.cc
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/tests/sys/fs/fusefs/getattr.cc
==
--- stable/12/tests/sys/fs/fusefs/getattr.ccFri Jun 12 19:56:19 2020
(r362113)
+++ stable/12/tests/sys/fs/fusefs/getattr.ccFri Jun 12 20:11:25 2020
(r362114)
@@ -32,6 +32,8 @@
extern "C" {
#include
+
+#include
}
#include "mockfs.hh"
@@ -172,7 +174,10 @@ TEST_F(Getattr, enoent)
const char RELPATH[] = "some_file.txt";
struct stat sb;
const uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino, S_IFREG | 0644, 0, 1, 0, 0);
EXPECT_CALL(*m_mock, process(
ResultOf([](auto in) {
@@ -181,8 +186,15 @@ TEST_F(Getattr, enoent)
}, Eq(true)),
_)
).WillOnce(Invoke(ReturnErrno(ENOENT)));
+ // Since FUSE_GETATTR returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
+
EXPECT_NE(0, stat(FULLPATH, ));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
TEST_F(Getattr, ok)
Modified: stable/12/tests/sys/fs/fusefs/open.cc
==
--- stable/12/tests/sys/fs/fusefs/open.cc Fri Jun 12 19:56:19 2020
(r362113)
+++ stable/12/tests/sys/fs/fusefs/open.cc Fri Jun 12 20:11:25 2020
(r362114)
@@ -32,7 +32,9 @@
extern "C" {
#include
+
#include
+#include
}
#include "mockfs.hh"
@@ -105,7 +107,10 @@ TEST_F(Open, enoent)
const char FULLPATH[] = "mountpoint/some_file.txt";
const char RELPATH[] = "some_file.txt";
uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino, S_IFREG | 0644, 0, 1);
EXPECT_CALL(*m_mock, process(
ResultOf([=](auto in) {
@@ -114,8 +119,15 @@ TEST_F(Open, enoent)
}, Eq(true)),
_)
).WillOnce(Invoke(ReturnErrno(ENOENT)));
+ // Since FUSE_OPEN returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
+
ASSERT_EQ(-1, open(FULLPATH, O_RDONLY));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
/*
Modified: stable/12/tests/sys/fs/fusefs/opendir.cc
==
--- stable/12/tests/sys/fs/fusefs/opendir.ccFri Jun 12 19:56:19 2020
(r362113)
+++ stable/12/tests/sys/fs/fusefs/opendir.ccFri Jun 12 20:11:25 2020
(r362114)
@@ -32,7 +32,9 @@
extern "C" {
#include
+
#include
+#include
}
#include "mockfs.hh"
@@ -82,12 +84,21 @@ TEST_F(Opendir, enoent)
const char FULLPATH[] = "mountpoint/some_dir";
const char RELPATH[] = "some_dir";
uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino);
expect_opendir(ino, O_RDONLY, ReturnErrno(ENOENT));
+ // Since FUSE_OPENDIR returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
ASSERT_EQ(-1, open(FULLPATH, O_DIRECTORY));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
/*
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361562 - head/sys/geom/eli
Author: asomers
Date: Wed May 27 19:13:26 2020
New Revision: 361562
URL: https://svnweb.freebsd.org/changeset/base/361562
Log:
geli: fix a livelock during panic
During any kind of shutdown, kern_reboot calls geli's pre_sync event hook,
which tries to destroy all unused geli devices. But during a panic, geli
can't destroy any devices, because the scheduler is stopped, so it can't
switch threads. A livelock results, and the system never dumps core.
This commit fixes the problem by refusing to destroy any devices during
panic, used or otherwise.
PR: 246207
Reviewed by: jhb
MFC after:2 weeks
Sponsored by: Axcient
Differential Revision:https://reviews.freebsd.org/D24697
Modified:
head/sys/geom/eli/g_eli.c
Modified: head/sys/geom/eli/g_eli.c
==
--- head/sys/geom/eli/g_eli.c Wed May 27 18:55:24 2020(r361561)
+++ head/sys/geom/eli/g_eli.c Wed May 27 19:13:26 2020(r361562)
@@ -1416,11 +1416,13 @@ g_eli_shutdown_pre_sync(void *arg, int howto)
continue;
pp = LIST_FIRST(>provider);
KASSERT(pp != NULL, ("No provider? gp=%p (%s)", gp, gp->name));
- if (pp->acr + pp->acw + pp->ace == 0)
- error = g_eli_destroy(sc, TRUE);
- else {
+ if (pp->acr != 0 || pp->acw != 0 || pp->ace != 0 ||
+ SCHEDULER_STOPPED())
+ {
sc->sc_flags |= G_ELI_FLAG_RW_DETACH;
gp->access = g_eli_access;
+ } else {
+ error = g_eli_destroy(sc, TRUE);
}
}
g_topology_unlock();
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361439 - head/share/man/man4
Author: asomers Date: Mon May 25 04:17:01 2020 New Revision: 361439 URL: https://svnweb.freebsd.org/changeset/base/361439 Log: [skip ci] ip.4: fix typos MFC after:2 weeks Modified: head/share/man/man4/ip.4 Modified: head/share/man/man4/ip.4 == --- head/share/man/man4/ip.4Sun May 24 21:42:47 2020(r361438) +++ head/share/man/man4/ip.4Mon May 25 04:17:01 2020(r361439) @@ -28,7 +28,7 @@ .\" @(#)ip.4 8.2 (Berkeley) 11/30/93 .\" $FreeBSD$ .\" -.Dd February 22, 2019 +.Dd May 24, 2020 .Dt IP 4 .Os .Sh NAME @@ -144,7 +144,7 @@ the .Xr recvmsg 2 call will return the destination .Tn IP -address and destination port or a +address and destination port for a .Tn UDP datagram. The @@ -155,13 +155,13 @@ structure points to a buffer that contains a .Vt cmsghdr structure followed by the -.Tn in_sockkaddr -structre. +.Tn sockaddr_in +structure. The .Vt cmsghdr fields have the following values: .Bd -literal -cmsg_len = CMSG_LEN(sizeof(struct in_sockaddr)) +cmsg_len = CMSG_LEN(sizeof(struct sockaddr_in)) cmsg_level = IPPROTO_IP cmsg_type = IP_ORIGDSTADDR .Ed ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361407 - stable/12/tests/sys/geom/class/multipath
Author: asomers
Date: Fri May 22 22:17:44 2020
New Revision: 361407
URL: https://svnweb.freebsd.org/changeset/base/361407
Log:
Reenable sys.geom.class.multipath.misc.fail_on_error on stable/12
The failing test was fixed by r361403. Direct commit to stable/12 because
the test was never disabled on head.
PR: 244158
Reported by: lwhsu
Modified:
stable/12/tests/sys/geom/class/multipath/misc.sh
Modified: stable/12/tests/sys/geom/class/multipath/misc.sh
==
--- stable/12/tests/sys/geom/class/multipath/misc.shFri May 22 22:13:55
2020(r361406)
+++ stable/12/tests/sys/geom/class/multipath/misc.shFri May 22 22:17:44
2020(r361407)
@@ -176,10 +176,6 @@ fail_on_error_head()
}
fail_on_error_body()
{
- if [ "$(atf_config_get ci false)" = "true" ]; then
- atf_skip "https://bugs.freebsd.org/244158;
- fi
-
load_gnop
load_gmultipath
md0=$(alloc_md)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361406 - stable/12/tests/sys/fs/fusefs
Author: asomers Date: Fri May 22 22:13:55 2020 New Revision: 361406 URL: https://svnweb.freebsd.org/changeset/base/361406 Log: MFC r360829: fusefs: fix two small bugs in the tests' expectations These two errors have been present since the tests' introduction. Coincidentally every test (I think there's only one) that cares about that field also works when the field's value is 0. Modified: stable/12/tests/sys/fs/fusefs/default_permissions.cc stable/12/tests/sys/fs/fusefs/default_permissions_privileged.cc Directory Properties: stable/12/ (props changed) Modified: stable/12/tests/sys/fs/fusefs/default_permissions.cc == --- stable/12/tests/sys/fs/fusefs/default_permissions.ccFri May 22 22:12:07 2020(r361405) +++ stable/12/tests/sys/fs/fusefs/default_permissions.ccFri May 22 22:13:55 2020(r361406) @@ -125,7 +125,7 @@ void expect_getattr(uint64_t ino, mode_t mode, uint64_ out.body.attr.attr.mode = mode; out.body.attr.attr.size = 0; out.body.attr.attr.uid = uid; - out.body.attr.attr.uid = gid; + out.body.attr.attr.gid = gid; out.body.attr.attr_valid = attr_valid; }))); } Modified: stable/12/tests/sys/fs/fusefs/default_permissions_privileged.cc == --- stable/12/tests/sys/fs/fusefs/default_permissions_privileged.cc Fri May 22 22:12:07 2020(r361405) +++ stable/12/tests/sys/fs/fusefs/default_permissions_privileged.cc Fri May 22 22:13:55 2020(r361406) @@ -85,7 +85,7 @@ void expect_getattr(uint64_t ino, mode_t mode, uint64_ out.body.attr.attr.mode = mode; out.body.attr.attr.size = 0; out.body.attr.attr.uid = uid; - out.body.attr.attr.uid = gid; + out.body.attr.attr.gid = gid; out.body.attr.attr_valid = attr_valid; }))); } ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361405 - stable/12/sys/fs/fuse
Author: asomers
Date: Fri May 22 22:12:07 2020
New Revision: 361405
URL: https://svnweb.freebsd.org/changeset/base/361405
Log:
MFC r360828:
fusefs: better dtrace probes for asynchronous invalidation operations
Modified:
stable/12/sys/fs/fuse/fuse_internal.c
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/sys/fs/fuse/fuse_internal.c
==
--- stable/12/sys/fs/fuse/fuse_internal.c Fri May 22 20:52:36 2020
(r361404)
+++ stable/12/sys/fs/fuse/fuse_internal.c Fri May 22 22:12:07 2020
(r361405)
@@ -377,8 +377,8 @@ fuse_internal_fsync(struct vnode *vp,
}
/* Asynchronous invalidation */
-SDT_PROBE_DEFINE2(fusefs, , internal, invalidate_cache_hit,
- "struct vnode*", "struct vnode*");
+SDT_PROBE_DEFINE3(fusefs, , internal, invalidate_entry,
+ "struct vnode*", "struct fuse_notify_inval_entry_out*", "char*");
int
fuse_internal_invalidate_entry(struct mount *mp, struct uio *uio)
{
@@ -407,6 +407,7 @@ fuse_internal_invalidate_entry(struct mount *mp, struc
else
err = fuse_internal_get_cached_vnode( mp, fnieo.parent,
LK_SHARED, );
+ SDT_PROBE3(fusefs, , internal, invalidate_entry, dvp, , name);
/*
* If dvp is not in the cache, then it must've been reclaimed. And
* since fuse_vnop_reclaim does a cache_purge, name's entry must've
@@ -435,6 +436,8 @@ fuse_internal_invalidate_entry(struct mount *mp, struc
return (0);
}
+SDT_PROBE_DEFINE2(fusefs, , internal, invalidate_inode,
+ "struct vnode*", "struct fuse_notify_inval_inode_out *");
int
fuse_internal_invalidate_inode(struct mount *mp, struct uio *uio)
{
@@ -450,6 +453,7 @@ fuse_internal_invalidate_inode(struct mount *mp, struc
else
err = fuse_internal_get_cached_vnode(mp, fniio.ino, LK_SHARED,
);
+ SDT_PROBE2(fusefs, , internal, invalidate_inode, vp, );
if (err != 0 || vp == NULL)
return (err);
/*
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361403 - stable/12/tests/sys/geom/class/multipath
Author: asomers
Date: Fri May 22 19:09:43 2020
New Revision: 361403
URL: https://svnweb.freebsd.org/changeset/base/361403
Log:
MFC r360807:
Fix the sys.geom.class.multipath.misc.fail_on_error test on stable/12
This test uses a gnop feature (delay probability) that isn't available on
stable/12. But it's unnecessary; the test works fine without it. Removing
it simplifies the test and, once MFCed, will allow it to pass on stable/12.
PR: 244158
Reported by: lwhsu
Modified:
stable/12/tests/sys/geom/class/multipath/misc.sh
Directory Properties:
stable/12/ (props changed)
Modified: stable/12/tests/sys/geom/class/multipath/misc.sh
==
--- stable/12/tests/sys/geom/class/multipath/misc.shFri May 22 18:54:56
2020(r361402)
+++ stable/12/tests/sys/geom/class/multipath/misc.shFri May 22 19:09:43
2020(r361403)
@@ -190,7 +190,7 @@ fail_on_error_body()
atf_check -s exit:0 gmultipath create "$name" ${md0}.nop ${md1}.nop
# The first I/O to the first path should fail, causing gmultipath to
# fail over to the second path.
- atf_check gnop configure -q 100 -r 100 -w 100 -x 100 ${md0}.nop
+ atf_check gnop configure -r 100 -w 100 ${md0}.nop
atf_check -s exit:0 -o ignore -e ignore dd if=/dev/zero
of=/dev/multipath/"$name" bs=4096 count=1
check_multipath_state ${md1}.nop "DEGRADED" "FAIL" "ACTIVE"
}
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361401 - in head: sys/fs/fuse tests/sys/fs/fusefs
Author: asomers
Date: Fri May 22 18:11:17 2020
New Revision: 361401
URL: https://svnweb.freebsd.org/changeset/base/361401
Log:
Fix issues with FUSE_ACCESS when default_permissions is disabled
This patch fixes two issues relating to FUSE_ACCESS when the
default_permissions mount option is disabled:
* VOP_ACCESS() calls with VADMIN set should never be sent to a fuse server
in the form of FUSE_ACCESS operations. The FUSE protocol has no equivalent
of VADMIN, so we must evaluate such things kernel-side, regardless of the
default_permissions setting.
* The FUSE protocol only requires FUSE_ACCESS to be sent for two purposes:
for the access(2) syscall and to check directory permissions for
searchability during lookup. FreeBSD sends it much more frequently, due to
differences between our VFS and Linux's, for which FUSE was designed. But
this patch does eliminate several cases not required by the FUSE protocol:
* for any FUSE_*XATTR operation
* when creating a new file
* when deleting a file
* when setting timestamps, such as by utimensat(2).
* Additionally, when default_permissions is disabled, this patch removes one
FUSE_GETATTR operation when deleting a file.
PR: 245689
Reported by: MooseFS FreeBSD Team
Reviewed by: cem
MFC after:2 weeks
Differential Revision:https://reviews.freebsd.org/D24777
Modified:
head/sys/fs/fuse/fuse_internal.c
head/sys/fs/fuse/fuse_vnops.c
head/tests/sys/fs/fusefs/access.cc
head/tests/sys/fs/fusefs/rename.cc
head/tests/sys/fs/fusefs/rmdir.cc
head/tests/sys/fs/fusefs/unlink.cc
head/tests/sys/fs/fusefs/utils.cc
head/tests/sys/fs/fusefs/utils.hh
head/tests/sys/fs/fusefs/xattr.cc
Modified: head/sys/fs/fuse/fuse_internal.c
==
--- head/sys/fs/fuse/fuse_internal.cFri May 22 18:10:46 2020
(r361400)
+++ head/sys/fs/fuse/fuse_internal.cFri May 22 18:11:17 2020
(r361401)
@@ -158,6 +158,7 @@ fuse_internal_get_cached_vnode(struct mount* mp, ino_t
return 0;
}
+SDT_PROBE_DEFINE0(fusefs, , internal, access_vadmin);
/* Synchronously send a FUSE_ACCESS operation */
int
fuse_internal_access(struct vnode *vp,
@@ -210,10 +211,18 @@ fuse_internal_access(struct vnode *vp,
va.va_gid, mode, cred, NULL);
}
+ if (mode & VADMIN) {
+ /*
+* The FUSE protocol doesn't have an equivalent of VADMIN, so
+* it's a bug if we ever reach this point with that bit set.
+*/
+ SDT_PROBE0(fusefs, , internal, access_vadmin);
+ }
+
if (!fsess_isimpl(mp, FUSE_ACCESS))
return 0;
- if ((mode & (VWRITE | VAPPEND | VADMIN)) != 0)
+ if ((mode & (VWRITE | VAPPEND)) != 0)
mask |= W_OK;
if ((mode & VREAD) != 0)
mask |= R_OK;
Modified: head/sys/fs/fuse/fuse_vnops.c
==
--- head/sys/fs/fuse/fuse_vnops.c Fri May 22 18:10:46 2020
(r361400)
+++ head/sys/fs/fuse/fuse_vnops.c Fri May 22 18:11:17 2020
(r361401)
@@ -235,6 +235,7 @@ fuse_extattr_check_cred(struct vnode *vp, int ns, stru
{
struct mount *mp = vnode_mount(vp);
struct fuse_data *data = fuse_get_mpdata(mp);
+ int default_permissions = data->dataflags & FSESS_DEFAULT_PERMISSIONS;
/*
* Kernel-invoked always succeeds.
@@ -248,12 +249,15 @@ fuse_extattr_check_cred(struct vnode *vp, int ns, stru
*/
switch (ns) {
case EXTATTR_NAMESPACE_SYSTEM:
- if (data->dataflags & FSESS_DEFAULT_PERMISSIONS) {
+ if (default_permissions) {
return (priv_check_cred(cred, PRIV_VFS_EXTATTR_SYSTEM));
}
- /* FALLTHROUGH */
+ return (0);
case EXTATTR_NAMESPACE_USER:
- return (fuse_internal_access(vp, accmode, td, cred));
+ if (default_permissions) {
+ return (fuse_internal_access(vp, accmode, td, cred));
+ }
+ return (0);
default:
return (EPERM);
}
@@ -985,6 +989,8 @@ fuse_vnop_lookup(struct vop_lookup_args *ap)
int wantparent = flags & (LOCKPARENT | WANTPARENT);
int islastcn = flags & ISLASTCN;
struct mount *mp = vnode_mount(dvp);
+ struct fuse_data *data = fuse_get_mpdata(mp);
+ int default_permissions = data->dataflags & FSESS_DEFAULT_PERMISSIONS;
int err = 0;
int lookup_err = 0;
@@ -1108,7 +1114,11 @@ fuse_vnop_lookup(struct vop_lookup_args *ap)
if (lookup_err) {
/* Entry not found */
if ((nameiop == CREATE || nameiop == RENAME) && islastcn) {
- err = fuse_internal_access(dvp, VWRITE,
svn commit: r361399 - head/sys/fs/fuse
Author: asomers Date: Fri May 22 18:03:14 2020 New Revision: 361399 URL: https://svnweb.freebsd.org/changeset/base/361399 Log: Disable nullfs cacheing on top of fusefs Nullfs cacheing can keep a large number of vnodes active. That results in more active FUSE file handles, causing some FUSE servers to use extra resources. Disable nullfs cacheing for fusefs, just like we already do for NFSv4. PR: 245688 Reported by: MooseFS FreeBSD Team MFC after:2 weeks Modified: head/sys/fs/fuse/fuse_vfsops.c Modified: head/sys/fs/fuse/fuse_vfsops.c == --- head/sys/fs/fuse/fuse_vfsops.c Fri May 22 17:52:09 2020 (r361398) +++ head/sys/fs/fuse/fuse_vfsops.c Fri May 22 18:03:14 2020 (r361399) @@ -425,6 +425,11 @@ fuse_vfsop_mount(struct mount *mp) */ mp->mnt_flag &= ~MNT_LOCAL; mp->mnt_kern_flag |= MNTK_USES_BCACHE; + /* +* Disable nullfs cacheing because it can consume too many resources in +* the FUSE server. +*/ + mp->mnt_kern_flag |= MNTK_NULL_NOCACHE; MNT_IUNLOCK(mp); /* We need this here as this slot is used by getnewvnode() */ mp->mnt_stat.f_iosize = maxbcachebuf; ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361223 - head/tests/sys/fs/fusefs
Author: asomers
Date: Mon May 18 18:36:32 2020
New Revision: 361223
URL: https://svnweb.freebsd.org/changeset/base/361223
Log:
fusefs: fix intermittency in some ENOENT tests
When a FUSE operation other than LOOKUP returns ENOENT, the kernel will
reclaim that vnode, resuling in a FUSE_FORGET being sent a short while
later. Many of the ENOENT tests weren't expecting those FUSE_FORGET
operations. They usually passed by luck since FUSE_FORGET is often delayed.
This commit adds appropriate expectations.
MFC after:2 weeks
Modified:
head/tests/sys/fs/fusefs/getattr.cc
head/tests/sys/fs/fusefs/open.cc
head/tests/sys/fs/fusefs/opendir.cc
Modified: head/tests/sys/fs/fusefs/getattr.cc
==
--- head/tests/sys/fs/fusefs/getattr.cc Mon May 18 18:32:58 2020
(r361222)
+++ head/tests/sys/fs/fusefs/getattr.cc Mon May 18 18:36:32 2020
(r361223)
@@ -32,6 +32,8 @@
extern "C" {
#include
+
+#include
}
#include "mockfs.hh"
@@ -172,7 +174,10 @@ TEST_F(Getattr, enoent)
const char RELPATH[] = "some_file.txt";
struct stat sb;
const uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino, S_IFREG | 0644, 0, 1, 0, 0);
EXPECT_CALL(*m_mock, process(
ResultOf([](auto in) {
@@ -181,8 +186,15 @@ TEST_F(Getattr, enoent)
}, Eq(true)),
_)
).WillOnce(Invoke(ReturnErrno(ENOENT)));
+ // Since FUSE_GETATTR returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
+
EXPECT_NE(0, stat(FULLPATH, ));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
TEST_F(Getattr, ok)
Modified: head/tests/sys/fs/fusefs/open.cc
==
--- head/tests/sys/fs/fusefs/open.ccMon May 18 18:32:58 2020
(r361222)
+++ head/tests/sys/fs/fusefs/open.ccMon May 18 18:36:32 2020
(r361223)
@@ -32,7 +32,9 @@
extern "C" {
#include
+
#include
+#include
}
#include "mockfs.hh"
@@ -105,7 +107,10 @@ TEST_F(Open, enoent)
const char FULLPATH[] = "mountpoint/some_file.txt";
const char RELPATH[] = "some_file.txt";
uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino, S_IFREG | 0644, 0, 1);
EXPECT_CALL(*m_mock, process(
ResultOf([=](auto in) {
@@ -114,8 +119,15 @@ TEST_F(Open, enoent)
}, Eq(true)),
_)
).WillOnce(Invoke(ReturnErrno(ENOENT)));
+ // Since FUSE_OPEN returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
+
ASSERT_EQ(-1, open(FULLPATH, O_RDONLY));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
/*
Modified: head/tests/sys/fs/fusefs/opendir.cc
==
--- head/tests/sys/fs/fusefs/opendir.cc Mon May 18 18:32:58 2020
(r361222)
+++ head/tests/sys/fs/fusefs/opendir.cc Mon May 18 18:36:32 2020
(r361223)
@@ -32,7 +32,9 @@
extern "C" {
#include
+
#include
+#include
}
#include "mockfs.hh"
@@ -82,12 +84,21 @@ TEST_F(Opendir, enoent)
const char FULLPATH[] = "mountpoint/some_dir";
const char RELPATH[] = "some_dir";
uint64_t ino = 42;
+ sem_t sem;
+ ASSERT_EQ(0, sem_init(, 0, 0)) << strerror(errno);
+
expect_lookup(RELPATH, ino);
expect_opendir(ino, O_RDONLY, ReturnErrno(ENOENT));
+ // Since FUSE_OPENDIR returns ENOENT, the kernel will reclaim the vnode
+ // and send a FUSE_FORGET
+ expect_forget(ino, 1, );
ASSERT_EQ(-1, open(FULLPATH, O_DIRECTORY));
EXPECT_EQ(ENOENT, errno);
+
+ sem_wait();
+ sem_destroy();
}
/*
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361129 - head/tests/sys/geom/class/gate
Author: asomers
Date: Sun May 17 02:41:50 2020
New Revision: 361129
URL: https://svnweb.freebsd.org/changeset/base/361129
Log:
Reenable sys.geom.class.gate.ggate_test.ggated in CI
Should be fixed by r360613
PR: 244737
Reported by: lwhsu
Modified:
head/tests/sys/geom/class/gate/ggate_test.sh
Modified: head/tests/sys/geom/class/gate/ggate_test.sh
==
--- head/tests/sys/geom/class/gate/ggate_test.shSun May 17 02:40:49
2020(r361128)
+++ head/tests/sys/geom/class/gate/ggate_test.shSun May 17 02:41:50
2020(r361129)
@@ -16,10 +16,6 @@ ggated_head()
ggated_body()
{
- if [ "$(atf_config_get ci false)" = "true" ]; then
- atf_skip "https://bugs.freebsd.org/244737;
- fi
-
load_ggate
us=$(alloc_ggate_dev)
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
svn commit: r361128 - stable/11/tests/sys/geom/class/gate
Author: asomers Date: Sun May 17 02:40:49 2020 New Revision: 361128 URL: https://svnweb.freebsd.org/changeset/base/361128 Log: MFC r360613: Fix intermittent cleanup failures in the ggated test Modified: stable/11/tests/sys/geom/class/gate/ggate_test.sh Directory Properties: stable/11/ (props changed) Modified: stable/11/tests/sys/geom/class/gate/ggate_test.sh == --- stable/11/tests/sys/geom/class/gate/ggate_test.sh Sun May 17 02:36:42 2020(r361127) +++ stable/11/tests/sys/geom/class/gate/ggate_test.sh Sun May 17 02:40:49 2020(r361128) @@ -194,7 +194,11 @@ common_cleanup() if [ -f "md.devs" ]; then while read test_md; do - mdconfig -d -u $test_md 2>/dev/null + # ggatec destroy doesn't release the provider + # synchronously, so we may need to retry destroying it. + while ! mdconfig -d -u $test_md; do + sleep 0.1 + done done < md.devs rm md.devs fi ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
