| Age | Commit message (Collapse) | Author |
|---|
|
- fix TimeToStr i18n (LP: #289807)
* [ABI break] merge support for http redirects, thanks to
Jeff Licquia and Anthony Towns
* [ABI break] use int for the package IDs (thanks to Steve Cotton)
* apt-pkg/pkgcache.cc:
- do not run "dpkg --configure pkg" if pkg is in trigger-awaited
state (LP: #322955)
* methods/https.cc:
- add Acquire::https::AllowRedirect support
* Clarify the --help for 'purge' (LP: #243948)
* cmdline/apt-get.cc
- fix "apt-get source pkg" if there is a binary package and
a source package of the same name but from different
packages (LP: #330103)
* cmdline/acqprogress.cc:
- Call pkgAcquireStatus::Pulse even if quiet, so that we still get
dlstatus messages on the status-fd (LP: #290234).
|
|
|
|
|
|
|
|
|
|
- move the state file writting into the Go() implementation
of dpkgpm (closes: #498799)
* apt-pkg/algorithms.cc:
- fix simulation performance drop (thanks to Ferenc Wagner
for reporting the issue)
|
|
The problem was that the size of pkgDpkgPM and its member offsets
changed because a map giving the names of the trigger states was
inserted into the middle of the structure. I fixed it by using a
statically allocated array instead. This changes the procedure for
looking up a string to a linear search, which should be fine (or
even faster than before) since there are only 4 state strings. If
it becomes a problem, sorting the array by key will allow us to use
std::equal_range(), but I would advise against this unless it's
really necessary, since sooner or later someone will forget to maintain
the sort order.
|
|
apt/aptitude (like the installer) defer trigger processing
(thanks to Joey Hess)
|
|
|
|
- merged patch from Kees Cook to fix anoying upper-case display
on amd64 in sbuild
|
|
for apt-get update like operations for the frontends and also provides
hooks to run stuff in APT::Update::{Pre,Post}-Invoke
|
|
|
|
getting crazy when /dev/null is redirected to stdin (which breaks
buildds), closes: #452858.
|
|
|
|
to add backward support for arches that lacks pselect support,
closes: #448406.
|
|
|
|
- use pselect() instead of select()
- on EIO error on pty master read, usleep(0.5s) to give up timeslice
so that the child can properly exit
|
|
apt-get check is successfull)
|
|
- fix resource leak
|
|
|
|
string constant to 'char*';
|
|
- make it compileable
|
|
|
|
- ignore EINTR on select()
|
|
- comment out dpkg trigger messages, those seem to be not generated by
dpkg over the status fd
|
|
|
|
- remove leading "\n"
|
|
|
|
|
|
- on error in select() continue
|
|
move term_out into DPkgPM class
|
|
|
|
- do not fail if openpty() fails
|
|
|
|
- init Dir::Log::Terminal
* apt-pkg/deb/dpkgpm.cc:
- only work with term_out if it is not NULL
|
|
- updated Dir::Log::Terminal
* apt-pkg/deb/dpkgpm.cc:
- make the terminal log optional
- rename Dir::Log::Name to Dir::Log::Terminal
|
|
- a bit cleanup
- move the log date to the right place
- write log to dir::log::name
apt-pkg/init.cc:
- init dir::log::name "/var/log/apt/term.log"
debian/apt.dirs:
- create /var/log/apt/
doc/examples/configure-index:
- add new dir::log::name to the index
|
|
- move output processing ProcessDpkgStatusLine() and out of the
Go() method (yeah!) - things start to look more tidy now
|
|
- move make dpkgstatus processing into DoDpkgStatusFd()
and ProcessDpkgStatusLine() [the later is a stub for now]
|
|
- merge select() based terminal log
|
|
- run DoStdin(), DoTerminalPty() calls only if their FD is in the
returned set after select()
|
|
- write current time to log
|
|
- unbreak progress reporting, make code nicer
|
|
- move the terminal and stdin reading into its own functions
|
|
- use select() to check for new data
|
|
|
|
|
|
|
|
|
|
- added "Dpkg::StopOnError" variable that controls if apt
will abort on errors from dpkg
|
