| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
|
|
|
| |
and also for pfSense-upgrade
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
(cherry picked from commit 05bdceee8c8fd598cdf83dc57c4006e20612e26d)
|
| |
|
|
|
|
| |
(cherry picked from commit 75cdfae39bff2b5ccc523cc32df7dde99f04eeed)
|
|
|
|
| |
correct one
|
|
|
|
|
|
| |
- Distribute only a single pfSense-repo package containing all templates
- Create a symlink pointing to selected repo
- Do not limit it only to pfSense-repo and pfSense-repo-devel
|
| |
|
|
|
|
|
|
| |
- Do not call ntpdate before start ntpd, ntpd -g parameter is enough
- Deprecate /usr/local/sbin/ntpdate_sync_once.sh
- Remove system_ntp_configure parameter and always start ntpd
|
| |
|
|
|
|
| |
changes. Fixes #6052
|
|
|
|
| |
specified. Remove old and wrongly copy/pasted comments while here. Ticket #5981
|
|
|
| |
Fix missing check for $active to avoid creating a route to a released lease.
|
|
|
|
| |
Made a silly mistake and overlooked an important difference between javascript substring() and php substr().
Tightened regex check for octal value.
|
|
|
|
| |
fixes #5875
|
|
|
|
|
|
| |
default gateway switching. Fixes #1837
Tested with default gateway switching on and off, by rebooting, manually disconnecting the PPP link, and manually disconnecting the NIC link, worked each time, though still possible some issues remain. Feedback appreciated.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since the introduction of pkg_with_pb(), direct calls to this function
were added using the following format:
if ! pkg_with_pb ... | tee -a $logfile; then
What happened in this case is if was always getting a successfull
return code from tee call, and pkg_with_pb return was always being
ignored leading pfSense-upgrade to keep running when a fatal error
happened.
Since we need to print pkg output and also send it to log, _exec() call
is the best option here.
Added necessary flags to make sure necessary actions (unlock additional
packages or cleaning up kernel package annotation) are executed on
_exit call in fatal cases and replaced all direct calls by _exec()
|
|
|
|
| |
it's booting
|
|
|
|
| |
This reverts commit 93a8ddf7a4c5d43cbeb623c923eb76659e6f49e0.
|
|
|
|
| |
returning 0. It should fix #5837
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
After PKG_AUTOUPDATE=false was removed, operations done during boot,
2nd and 3rd stages of upgrade process, are trying to update repository
metadata. The problem is during this stage we don't have a network
available yet and all needed packages were previously downloaded.
Add -U parameter to all read/write pkg calls during at these stages to
prevent pkg(8) to try to update metadata
|
|
|
|
| |
times in some circumstances. Ticket #5621
|
| |
|
|
|
|
|
| |
After base64 encoding username/password, properly escape characters =,+,/
before submitting auth details
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
This patch addresses:
1. Handling of IA_NA and IA_PD strings (that contain IAID+DUID content) using only the DUID part.
2. Fixing regular expression matching with respect to the IAID+DUID string regarding the legal \" substring (used in ISC DHCPv6 leases).
3. Checking the $duid variable before use. Default case for $type in the switch case statement.
Please see the ticket for further information.
|
| |
|
| |
|
|
|
|
|
| |
* 'bugfix-for-4605' of https://github.com/the-real-ed/pfsense:
Bug fix for #4605. After base64 encoding username/password, properly escape characters =,+,/ before submitting auth details
|
|
|
|
| |
it's necessary. Some users reported to see issues in cases where update was necessary
|
|
|
|
| |
was the reason they were added, it was never finished and it's not being used
|
|
|
|
|
|
|
|
|
|
|
| |
On nanoBSD there is a "Proceed with upgrade?" question warning about the duplicate slice. After answering "y" to that, the system does the duplicate slice, which takes some minutes. Then it asks again "Proceed with upgrade?" after displaying all the packages it will install.
That is a bit annoying at the console - I answer "y" and go off to make a cup of tea, only to come back and find that it is waiting asking again.
On nanoBSD a reboot is known to be required anyway, because even if it is some little package that gets upgraded, and not the core OS, the partition is always switched. So we can say that in the first warning and then skip asking "Proceed with upgrade?" a second time.
What do you think? Should it ask a 2nd time after displaying the packages to be installed? Or is it OK to skip that confirmation prompt?
Also, I moved the call to setup_nanobsd_env inside the "if nanobsd" - it worked like it was because setup_nanobsd_env returns without doing anything if the system is not nanobsd, but it just looked odd and there seems no point calling it when not nanobsd.
|
|
|
|
|
|
| |
When -p is used on nanobsd, progres socket must be created on secondary
partition (chroot_dir) otherwise pkg, that runs with -c chroot_dir,
will not be able to find UNIX socket to connect and send events
|
| |
|
|
|
| |
Just a few extra words to put in this text, while I am reading it and waiting for the upgrade to complete.
|
| |
|