Remove OSVERSION checks that do not make sense any more.
For example (${OSVERSION} >= 900000 && ${OSVERSION} < 900021) is always true,
as is (${OSVERSION} > 900002 || ${OSVERSION} < 900000 && ${OSVERSION} > 800107).
Regarding patches, when an EXTRA_PATCHES is no longer needed, I remove it, when
it is always needed, I renamed it, in one case, I merged two patches.
Differential Revision: https://reviews.freebsd.org/D2209
sysutils/fusefs-kmod: Add PORTEPOCH
For a reason unexplained in the PR, the version went backwards on the
previous update. I guess this went unnoticed until now, but setting the
PORTEPOCH now is better late than never.
PR: 184042
Convert to USES=kmod, which removes duplicated code and ensures that all
required steps are followed.
There are no user-visible changes. The exception is multimedia/ptx-kmod,
which now installs the kernel module into /boot/modules instead of
${PREFIX}/libexec/${PORTNAME}, this was something USES=kmod cannot handle.
PR: ports/183625
Submitted by: myself
Approved by: portmgr (bdrewery)
Exp-run by: bdrewery
SSP support has been added to ports with WITH_SSP for i386 and amd64
on FreeBSD 10, and amd64 on earlier versions.
SSP_UNSAFE is added to disable in a port if it fails to build, but
this should only be used in rare circumstances such as kernel modules.
Otherwise, the port may just be failing due to lack of respecting
LDFLAGS.
On FreeBSD 10, this uses an ldscript in /usr/lib/libc.so to pull in
libssp_nonshared.a to address issues linking on i386 [1].
On earlier FreeBSD versions the WITH_SSP knob will add -lssp_nonshared
to LDFLAGS on i386. This is not needed on amd64. However, several hundred
ports do not currently respect LDFLAGS, so this support is disabled currently
as it causes build failures if a dependency is looking for the stack_chk
(Only the first 15 lines of the commit message are shown above )
- update sysutils/fusefs-libs to 2.9.1
- add a patch to sysutils/fusefs-kmod to make it compile with new fusefs-libs
Tested by: pho, Kevin Oberman <kob6558@gmail.com>,
Gustau Perez Querol <gperez@entel.upc.edu>
- re-order, include bsd.port.pre.mk so SRC_BASE is defined before referenced
- pet portlint on OPTIONS order
PR: ports/169221
Submitted by: jgh@
Approved by: maintainer timeout ( 14 days )
- The mount checkpoint was changed in previous update to r230226.
- This patch fixed mount_fusefs to use new path.
- Bump PORTREVISION
PR: ports/164218
Submitted by: jh@
Approved by: Edward Sanford Sutton, III <mirror176@cox.net> (maintainer),
gabor (mentor)
In the rc.d scripts, change assignments to rcvar to use the
literal name_enable wherever possible, and ${name}_enable
when it's not, to prepare for the demise of set_rcvar().
In cases where I had to hand-edit unusual instances also
modify formatting slightly to be more uniform (and in
some cases, correct). This includes adding some $FreeBSD$
tags, and most importantly moving rcvar= to right after
name= so it's clear that one is derived from the other.
- last submitter has become maintainer, in private email
- "Edward Sanford Sutton, III" <mirror176@cox.net>
- fusefs will be in 9.x anyway, and at that time, a patch to IGNORE 9.x will be
submitted
Submitted by: mirror176@cox.net
Approved by: maintainer (mirror176@cox.net), gabor (mentor)
Feature safe: yes
-Submitted with changes (no portbump per gabor)
-MNT_NFS4ACLS being defined in /usr/src/sys/mount.h but not being included in
the defined includes
-it is now included
PR: 151296
Submitted by: Edward.Sanford.Sutton, III <mirror176@cox.net>
Approved by: ports (maintainer timeout) gabor (mentor)
Feature safe: yes
Use __FreeBSD_version instead of __FreeBSD__ macro to compile on Debian
GNU/kFreeBSD
PR: ports/158825
Submitted by: Robert Millan <rmh@debian.org>
Approved by: maintainer timeout
- Mark as broken on 8.x. There are two PRs about this. The first (ports/135285)
has a patch but that causes coredumps. The second (ports/135621) has not
been tested yet but the port should still be marked as broken until a
proper fix is put in place.
PR: ports/135285
Submitted by: David Naylor <naylor.b.david@gmail.com>
Approved by: maintainer timeout
Fix build on -CURRENT after the removal of the thread argument from the FSD
(File-System Dependent) parts of the VFS.
Approved by: Anish Mistry (amistry at am-productions dot biz) (maintainer)
Make the fusefs kernel module compile on -CURRENT again.
The kernel module uses unit2minor(), which is not available on -CURRENT
anymore. I'm not increasing PORTREVISION, because this commit should not
have any effect on <800062 users.
Approved by: port maintainer
- Add .undef KERNCONF to fusefs-kmod's Makefile, so KERNCONF defined in
/etc/make.conf don't break the build in some cases
PR: 129909
Submitted by: myself
Approved by: Alex Kozlov <spam at rm-rf dot kiev dot ua> (maintainer)
Add knobs to the rc.d script to safely wait until all data has been written on
shutdown. (As this implies pausing rc.shutdown's shutdown watchdog, the knobs
are off by default.)
PR: 124901
Submitted by: Dominic Fandrey <kamikaze@bsdforen.de>
Approved by: maintainer
Fix build on -CURRENT after the introduction of accmode_t and renaming of
VOP_ACCESS 'a_mode' argument to 'a_accmode'.
Reviewed by: Anish Mistry <amistry at am-productions dot biz> (maintainer)
Distfile re-rolled. According to the maintainer all that changed was
some mercurial files were removed.
PR: ports/125734 [1], ports/125495 [2]
Submitted by: Anish Mistry <amistry@am-productions.biz> (maintainer) [1]
Eric J Christeson <Eric.J.Christeson@gmail.com> [2]
Make fusefs-kmod work again, after the addition of D_NEEDMINOR.
Drivers that use the clone_* KPI need to be marked with D_NEEDMINOR to
work right now. Without this patch, we'll get a kernel panic when
allocating the first device node.
Approved by: philip (mentor)
- When KERNCONF is set in /etc/make.conf and /usr/src is a symlink,
fuse's Makefile uses hardcoded /usr/obj/usr/src/sys/${KERNCONF}.
This causes build failure. Fix this by getting correct obj directory
from /usr/src/Makefile.inc1.
PR: ports/122257
Submitted by: rafan
Approved by: Anish Mistry <amistry at am-productions.biz> (maintainer)
From [1]:
- Update to version 0.3.9-pre1.20080208 (agreed with author).
- Download directly from the Mercurial reposirory (as tarball), to ease
maintainence.
- Remove all local patches as they were integrated.
- Convert setup.sh into setup.sh.in, and use SUB_FILES instead of manual
replacement.
- Use kldload instead of rc.d script in setup.sh.in because the latter gives
an error.
- Remove "sleep 1" from rc.d script as now the FUSE daemon makes umount wait
until disks are synced (synchronous unmount).
- Update pkg-message.
From [2]:
- Pass SRC_BASE to the actual build (previously was only used by the port to
detect if the source was installed, but not passed to the internal Makefile).
PR: ports/120420 [1], ports/118112 [2]
Submitted by: alepulver [1], Yuri Pankov <yuri.pankov@gmail.com> [2]
Approved by: maintainer (timeout) [1], maintainer [2]
- Fix build on 8-CURRENT. It was broken because of vn_lock/VOP_LOCK changes.
- Pet portlint, while here.
Approved by: maintainer timeout (more than 2 weeks)
Related PR: 119764
- Add support for SRC_BASE and MAKEOBJDIRPREFIX if using KERNCONF
PR: ports/115954
Submitted by: Nikos Ntarmos <ntarmos at ceid.upatras.gr>
Reworked by: mm
Approved by: maintainer timeout
- Bump PORTREVISION.
- Add checks in rc.d scripts to avoid "double" loading/unloading.
- Add patch-fuse_io.c (fix from
http://mercurial.creo.hu/repos/fuse4bsd-hg/?rev/d7bf90a8cc85).
- Handle "nonempty" option by ignoring it, used by ntfs-3g
(http://mercurial.creo.hu/repos/fuse4bsd-hg/?rev/aef999f12049).
- In the rc.d script unmount FUSE filesystem in the "stop" command, in reverse
order in case the mounts are nested, waiting one second after each. This is
a temporary workaround for umount returning before the cache has been
flushed (which causes for example files not being written to the disk), a
proper solution is being worked on.
- Add note in pkg-message about "stop" rc.d command unmounting FUSE
filesystems.
PR: ports/118758
Submitted by: alepulver (myself)
Approved by: Anish Mistry <amistry@am-productions.biz> (maintainer)
- Bump PORTREVISION.
- Create pkg-message from SUB_FILES instead of using ${SED} to print.
- Add a note in pkg-message and clean up.
- Create a symlink of ${PREFIX}/sbin/mount_fusefs to /usr/sbin to allow
mounting from /etc/fstab.
- Add "BEFORE: mountlate" to rc.d script.
This has been verified to work with sysutils/fusefs-ntfs (see changes to
pkg-message in last commit).
PR: ports/117516
Submitted by: alepulver (myself)
Approved by: Anish Mistry <amistry@am-productions.biz> (maintainer)
Sysutils/fusefs-kmod requires also userland sources
Fusefs-kmod says it requires kernel sources, but it also requires
userland sources:
Warning: Object directory not changed from original
/var/ports/usr/ports/sysutils/fusefs-kmod/work/fuse4bsd-0.3.9-pre1/mount_fusefs
cc -O2 -fno-strict-aliasing -pipe -I/usr/src/sbin/mount -I../include -c
mount_fusefs.c
^^^^^^^^^^^^^^^^^^^^^^
mount_fusefs.c:50:21: mntopts.h: No such file or directory
mount_fusefs.c:57: error: elements of array `mopts' have incomplete type
PR: ports/116196
Submitted by: Edwin Groothuis <edwin@mavetju.org>
Approved by: maintainer timeout
- Add patch for better processing of KERNCONF variable on build time
PR: ports/115585
Submitted by: Marcin Wisnicki <mwisnicki+freebsd at gmail.com>
Approved by: maintainer
- Annotate sysctl.conf entry in rc script comment block for people who miss
pkg-message
PR: ports/107466
Submitted by: Peter Risdon <prisdon@gmail.com>
Approved by: maintainer timeout
- Fix a bug in the initialization of a debug kernel sysctl. Merge changeset
9131f671c86d from upstream. This should fix problems with errors and
truncations during a save.
PR: ports/103722
Submitted by: Anish Mistry <amistry@am-productions.biz> (maintainer)
Remove the FreeBSD KEYWORD from all rc.d scripts where it appears.
We have not checked for this KEYWORD for a long time now, so this
is a complete noop, and thus no PORTREVISION bump. Removing it at
this point is mostly for pedantic reasons, and partly to avoid
perpetuating this anachronism by copy and paste to future scripts.
