]>
git.saurik.com Git - apt.git/log
David Kalnischkies [Mon, 19 Sep 2011 11:31:29 +0000 (13:31 +0200)]
do not pollute namespace in the headers with using (Closes: #500198)
David Kalnischkies [Mon, 19 Sep 2011 10:26:56 +0000 (12:26 +0200)]
remove an extra argument for the error mesage format
Michael Vogt [Wed, 14 Sep 2011 19:19:05 +0000 (21:19 +0200)]
releasing version 0.8.16~exp6
Michael Vogt [Wed, 14 Sep 2011 18:30:15 +0000 (20:30 +0200)]
merged from donkult
David Kalnischkies [Wed, 14 Sep 2011 18:22:31 +0000 (20:22 +0200)]
enable APT in unpack/configure ordering to handle loops as well
as tight dependencies between immediate packages better
enabling also the possibility to mark all packages as immediate
(at least Closes: #353290, #540227, #559733, #621836, #639290)
Michael Vogt [Wed, 14 Sep 2011 18:14:30 +0000 (20:14 +0200)]
merged from donkult
David Kalnischkies [Wed, 14 Sep 2011 13:48:18 +0000 (15:48 +0200)]
update symbols
Michael Vogt [Wed, 14 Sep 2011 11:56:33 +0000 (13:56 +0200)]
update package names too for the new abi
Michael Vogt [Wed, 14 Sep 2011 11:24:23 +0000 (13:24 +0200)]
* [abi-break] Support large files in the complete toolset. Indexes of this
* bump ABI version
Michael Vogt [Wed, 14 Sep 2011 11:22:19 +0000 (13:22 +0200)]
merged from lp:~donkult/apt/experimental
David Kalnischkies [Tue, 13 Sep 2011 16:00:22 +0000 (18:00 +0200)]
merge with debian/sid
David Kalnischkies [Tue, 13 Sep 2011 15:52:22 +0000 (17:52 +0200)]
merge with debian/experimental
David Kalnischkies [Tue, 13 Sep 2011 15:46:48 +0000 (17:46 +0200)]
Support large files in the complete toolset. Indexes of this
size are pretty unlikely for now, but we need it for deb
packages which could become bigger than 4GB now (LP: #815895)
David Kalnischkies [Tue, 13 Sep 2011 08:09:00 +0000 (10:09 +0200)]
reorder includes: add <config.h> if needed and include it at first
Michael Vogt [Mon, 12 Sep 2011 16:25:41 +0000 (18:25 +0200)]
releasing version 0.8.15.7
Michael Vogt [Mon, 12 Sep 2011 14:09:53 +0000 (16:09 +0200)]
merged from lp:~donkult/apt/sid
David Kalnischkies [Fri, 9 Sep 2011 10:51:18 +0000 (12:51 +0200)]
fix changelog typos and bugfix-link
David Kalnischkies [Fri, 9 Sep 2011 10:35:22 +0000 (12:35 +0200)]
M-A:same lockstep unpack should operate on installed
packages first (LP: #835625)
David Kalnischkies [Fri, 9 Sep 2011 08:56:42 +0000 (10:56 +0200)]
fix bashism (local outside function) and releasefile creation in testcases
David Kalnischkies [Mon, 5 Sep 2011 13:58:19 +0000 (15:58 +0200)]
- add an Acquire::Min-ValidTime option (Closes: #640122)
* doc/apt.conf.5.xml:
- reword Acquire::Max-ValidTime documentation to make clear
that it doesn't provide the new Min-ValidTime functionality
David Kalnischkies [Mon, 5 Sep 2011 12:50:30 +0000 (14:50 +0200)]
* apt-pkg/indexrecords.cc:
- fix Acquire::Max-ValidTime option by interpreting it really
as seconds as specified in the manpage and not as days
David Kalnischkies [Tue, 23 Aug 2011 22:41:30 +0000 (00:41 +0200)]
* apt-pkg/orderlist.cc:
- prefer visiting packages marked for deletion in VisitProvides
if we are operating on a negative dependency so that we can
deal early with the fallout of this remove
David Kalnischkies [Tue, 23 Aug 2011 18:17:19 +0000 (20:17 +0200)]
print from their the visit came from
David Kalnischkies [Tue, 23 Aug 2011 09:20:51 +0000 (11:20 +0200)]
* cmdline/apt-key:
- if command is 'add' do not error out if the specified
keyring doesn't exist, it will be created by gpg
David Kalnischkies [Mon, 22 Aug 2011 21:10:15 +0000 (23:10 +0200)]
remove the caches in 'apt-get update', too, as they will be
invalid in most cases anyway
David Kalnischkies [Mon, 22 Aug 2011 21:08:16 +0000 (23:08 +0200)]
* apt-pkg/acquire.cc:
- non-existing directories are by definition clean
David Kalnischkies [Mon, 22 Aug 2011 20:52:25 +0000 (22:52 +0200)]
* apt-pkg/acquire-item.cc:
- if no Release.gpg file is found try to verify with hashes,
but do not fail if a hash can't be found
David Kalnischkies [Thu, 18 Aug 2011 09:45:51 +0000 (11:45 +0200)]
* cmdline/apt-get.cc:
- remove the binary caches in 'apt-get clean' as it is the first
thing recommend by many supporters in case of APT segfaults
Michael Vogt [Wed, 17 Aug 2011 16:14:27 +0000 (18:14 +0200)]
merged lp:~donkult/apt/sid
David Kalnischkies [Wed, 17 Aug 2011 13:49:59 +0000 (15:49 +0200)]
set the msglevel to a high dummy value if none is specified
David Kalnischkies [Wed, 17 Aug 2011 13:47:08 +0000 (15:47 +0200)]
for consistency allow the usage of 'make test' instead of './run-tests'
David Kalnischkies [Wed, 17 Aug 2011 13:14:45 +0000 (15:14 +0200)]
understand DEB_BUILD_OPTIONS nocheck to disable the testexecution
David Kalnischkies [Wed, 17 Aug 2011 13:09:57 +0000 (15:09 +0200)]
* debian/apt.symbols:
- add the newly added symbols since 0.8.15.3
David Kalnischkies [Wed, 17 Aug 2011 12:49:03 +0000 (14:49 +0200)]
* test/*
- reorganize the various testcases and helper we have and
integrate them better into the buildsystem
- run the test/libapt testcases at package build-time
David Kalnischkies [Wed, 17 Aug 2011 12:41:02 +0000 (14:41 +0200)]
fix 'comparison between signed and unsigned integer expressions' warning
in the udevcdrom testcase
David Kalnischkies [Wed, 17 Aug 2011 12:37:25 +0000 (14:37 +0200)]
do not output color in the testcases if stdout/stderr are not a /dev/pts
David Kalnischkies [Wed, 17 Aug 2011 12:35:58 +0000 (14:35 +0200)]
build test-binaries by default and add a test target to toplevel
David Kalnischkies [Wed, 17 Aug 2011 10:39:59 +0000 (12:39 +0200)]
fix the compareversion test to work then called from outside of libapt
David Kalnischkies [Wed, 17 Aug 2011 10:15:39 +0000 (12:15 +0200)]
the experimental branch has a better version of this test in libapt,
so remove it now - we will get it back at merge-time with experimental
David Kalnischkies [Wed, 17 Aug 2011 10:15:21 +0000 (12:15 +0200)]
remove fake-status leftover from the python-test
David Kalnischkies [Wed, 17 Aug 2011 10:13:38 +0000 (12:13 +0200)]
move the networkless test to the interactive helpers as it needs
root rights for operation (ifup/down, iptables) so it is better run
under supervision
David Kalnischkies [Wed, 17 Aug 2011 10:09:12 +0000 (12:09 +0200)]
remove the python testcases as this is covered by integration/
David Kalnischkies [Wed, 17 Aug 2011 10:06:37 +0000 (12:06 +0200)]
move the interactive helper to a subdirectory of test
David Kalnischkies [Wed, 17 Aug 2011 10:05:22 +0000 (12:05 +0200)]
add 'dirs' target to the 'all' target to ensure that for e.g. the tests
all (sub)directories are build which are needed (e.g. obj/ and co)
David Kalnischkies [Wed, 17 Aug 2011 09:42:02 +0000 (11:42 +0200)]
move the list of versions to test with to the libapt subdirectory
Michael Vogt [Wed, 17 Aug 2011 09:38:45 +0000 (11:38 +0200)]
merged from lp:~donkult/apt/sid
David Kalnischkies [Wed, 17 Aug 2011 09:38:10 +0000 (11:38 +0200)]
remove the interactive scratch test, it doesn't test much and what
it tests is covered by your non-interactive testcases,
so this one wasn't used for a looooong time
David Kalnischkies [Wed, 17 Aug 2011 09:34:10 +0000 (11:34 +0200)]
replaced by the integration/test-compressed-indexes test
David Kalnischkies [Wed, 17 Aug 2011 09:15:36 +0000 (11:15 +0200)]
merge with the final changelog from debian-sid
David Kalnischkies [Wed, 17 Aug 2011 09:13:50 +0000 (11:13 +0200)]
* apt-pkg/packagemanager.cc, apt-pkg/pkgcache.cc:
- ignore "self"-conflicts for all architectures of a package
instead of just for the architecture of the package locked at
in the ordering of installations too (Closes: #802901)
David Kalnischkies [Wed, 17 Aug 2011 07:59:19 +0000 (09:59 +0200)]
merge the tests for configuration into another libapt-test
Christopher Baines [Tue, 16 Aug 2011 17:00:01 +0000 (18:00 +0100)]
Fixed a problem where the loop flag would be removed prematurely.
SmartConfigure xserver-xorg-video-apm (1:1.2.3-0ubuntu1)
SmartConfigure xserver-xorg-core (2:1.9.0-0ubuntu7.3) <- Loop flag set on xserver-xorg-core
SmartConfigure xserver-xorg (1:7.5+6ubuntu3)
SmartConfigure xserver-xorg-core (2:1.9.0-0ubuntu7.3) (Only Correct Dependancies) <- Loop flag removed prematurely
SmartConfigure libpciaccess0 (0.12.0-1)
SmartConfigure libpixman-1-0 (0.18.4-1)
SmartConfigure xserver-xorg-video-all (1:7.5+6ubuntu3)
SmartConfigure xserver-xorg-video-apm (1:1.2.3-0ubuntu1) (Only Correct Dependancies)
SmartConfigure xserver-xorg-core (2:1.9.0-0ubuntu7.3) <- Incorrectly detects first run as no loop flag
Also applied this fix to the SmartUnpack method.
Christopher Baines [Mon, 15 Aug 2011 21:31:09 +0000 (22:31 +0100)]
Fixed a bug where SmartUnPack would be called with Immediate=true (to resolve dependancies in SmartConfigure) yet Pkg
would not be immediately configured. This was because SmartUnPack still required the immediate flag to be set on Pkg.
Also inproved the debuging adding indented output for SmartUnPack and SmartConfigure and specifying in the output if
SmartConfigure or SmartUnPack was called just to Correct something (PkgLoop = true) or not.
David Kalnischkies [Mon, 15 Aug 2011 16:23:54 +0000 (18:23 +0200)]
support yet another webserver in the integration tests: gatling
David Kalnischkies [Mon, 15 Aug 2011 16:22:44 +0000 (18:22 +0200)]
write a proper testcase replacing the print-only uri.cc test
Michael Vogt [Mon, 15 Aug 2011 13:35:01 +0000 (15:35 +0200)]
releasing version 0.8.16~exp5
Michael Vogt [Mon, 15 Aug 2011 12:21:31 +0000 (14:21 +0200)]
merged from lp:~mvo/apt/mvo
Michael Vogt [Mon, 15 Aug 2011 12:10:42 +0000 (14:10 +0200)]
merged from the debian-sid branch
Michael Vogt [Mon, 15 Aug 2011 11:50:22 +0000 (13:50 +0200)]
releasing version 0.8.15.6
Michael Vogt [Mon, 15 Aug 2011 07:23:39 +0000 (09:23 +0200)]
merged fixes from lp:~mvo/apt/mvo
Christopher Baines [Sun, 14 Aug 2011 17:41:38 +0000 (18:41 +0100)]
Applied DonKult (David)'s excellent fix for inproving the loop management. Now both SmartConfigure and SmartUnPack can be
called mutiple times on the same package, this is to make sure that when loops are broken all packages that are required
are kept in the same dpkg run.
Christopher Baines [Sat, 13 Aug 2011 16:29:49 +0000 (17:29 +0100)]
Fix a bug introduced in Rev.2159 on line 398, also fix another potential bug.
Christopher Baines [Fri, 12 Aug 2011 15:36:25 +0000 (16:36 +0100)]
Added code to allow SmartConfigure to be called mutiple times on the same package to ensure all dependancies are satisfied.
Christopher Baines [Fri, 12 Aug 2011 11:22:17 +0000 (12:22 +0100)]
Inproved debug with versioning
Christopher Baines [Fri, 12 Aug 2011 09:38:19 +0000 (10:38 +0100)]
Small fix for loop handeling.
Michael Vogt [Fri, 12 Aug 2011 08:32:05 +0000 (10:32 +0200)]
lots of cppcheck fixes
Michael Vogt [Fri, 12 Aug 2011 07:58:55 +0000 (09:58 +0200)]
cherry pick test/integration/Packages-hashsum-verification addition from lp:~mvo/apt/mvo
Michael Vogt [Fri, 12 Aug 2011 07:56:14 +0000 (09:56 +0200)]
test/integration/Packages-hashsum-verification: commit the forgotten packages file
David Kalnischkies [Thu, 11 Aug 2011 21:30:09 +0000 (23:30 +0200)]
follow the recommendation of cppcheck to make some method methods (scnr)
const and initial mostly Debug member values in the constructors
David Kalnischkies [Thu, 11 Aug 2011 21:06:46 +0000 (23:06 +0200)]
rename the parameter name of MarkDelete from Purge to MarkPurge to
fix a cosmetic warning from cppcheck:
[apt-pkg/depcache.h:462] -> [apt-pkg/depcache.h:122]:
(style) Variable 'Purge' hides enumerator with same name
David Kalnischkies [Thu, 11 Aug 2011 21:01:15 +0000 (23:01 +0200)]
fix a few more cppcheck performance and scope warnings
David Kalnischkies [Thu, 11 Aug 2011 18:53:28 +0000 (20:53 +0200)]
cppcheck complains about some possible speed improvements which could be
done on the mirco-optimazation level, so lets fix them:
(performance) Possible inefficient checking for emptiness.
(performance) Prefer prefix ++/-- operators for non-primitive types.
David Kalnischkies [Thu, 11 Aug 2011 17:20:53 +0000 (19:20 +0200)]
fix some cppcheck: (warning) Member variable is not initialized in the constructor.
David Kalnischkies [Thu, 11 Aug 2011 16:42:19 +0000 (18:42 +0200)]
cppcheck is right that the check for preventing null deference of Query
is redundant in Redirect() as we can't reach the code with null anyway
[apt-pkg/acquire-method.cc:433]: (error) Possible null pointer dereference:
Queue - otherwise it is redundant to check if Queue is null at line 425
Christopher Baines [Thu, 11 Aug 2011 14:35:54 +0000 (15:35 +0100)]
Added a RmFlag function to remvoe the loop flag, this should prevent any errors or wierd behaviour because of the loop flag being used
at mutiple stages in both SmartUnpack and SmartConfigure.
Michael Vogt [Thu, 11 Aug 2011 14:30:17 +0000 (16:30 +0200)]
* methods/mirror.cc:
- include the architecture(s) in the query string as well so
that the server can make better decisions
Christopher Baines [Tue, 9 Aug 2011 16:10:09 +0000 (17:10 +0100)]
More inproved comments about loops.
Michael Vogt [Tue, 9 Aug 2011 12:50:20 +0000 (14:50 +0200)]
* apt-pkg/acquire.cc:
- fix potential divide-by-zero
Michael Vogt [Tue, 9 Aug 2011 12:38:01 +0000 (14:38 +0200)]
merged from the debian-experimental2 branch
Michael Vogt [Mon, 8 Aug 2011 14:18:52 +0000 (16:18 +0200)]
fix crash when P.Arch() was used but the cache got remapped
Michael Vogt [Mon, 8 Aug 2011 14:14:54 +0000 (16:14 +0200)]
apt-pkg/acquire-item.cc: add more debug output
Michael Vogt [Mon, 8 Aug 2011 14:14:13 +0000 (16:14 +0200)]
apt-pkg/pkgcachegen.cc: fix incorrect test for modification time
Christopher Baines [Sun, 7 Aug 2011 19:26:31 +0000 (20:26 +0100)]
Improved errors and messages in general and improved the comments.
Removed quite a bit of code I used while learning about how apt handles things.
Added some extra checks and warnings relevent for Immediate Configuration.
Removed a wierd section I put in to prevent a segfault at 724+, this appears no longer to be needed.
Christopher Baines [Sun, 7 Aug 2011 15:35:36 +0000 (16:35 +0100)]
Fixed a bug on line 623, I picked this up after seeing SmartUnpack trying to remove packages once they were confiured to solve Conflicts with the previous version!
Luckily EarlyRemove is sane, and properly checks, so I think this was just cosmetic.
Also fixed a bug on line 374 with SmartUnpack not checking if a dependancy has been removed, this bug was definately harmful.
Michael Vogt [Fri, 5 Aug 2011 10:29:40 +0000 (12:29 +0200)]
* test/integration/test-hashsum-verification:
- add regression test for hashsum verification
* apt-pkg/acquire-item.cc:
- if no Release.gpg file is found, still load the hashes for
verification (closes: #636314) and add test
Michael Vogt [Fri, 5 Aug 2011 10:26:35 +0000 (12:26 +0200)]
* apt-pkg/acquire-item.cc:
- if no Release.gpg file is found, still load the hashes for
verification (closes: #636314) and add test
Michael Vogt [Fri, 5 Aug 2011 09:49:59 +0000 (11:49 +0200)]
* test/integration/test-hashsum-verification:
- add regression test for hashsum verification
Michael Vogt [Fri, 5 Aug 2011 09:00:46 +0000 (11:00 +0200)]
apt-pkg/acquire-item.cc: always init Verify
Michael Vogt [Fri, 5 Aug 2011 08:48:18 +0000 (10:48 +0200)]
* apt-pkg/acquire-item.{cc,h}:
- do not check for a "Package" tag in optional index targets
like the translations index
Michael Vogt [Fri, 5 Aug 2011 07:12:31 +0000 (09:12 +0200)]
apt-pkg/pkgcachegen.cc: fix compiler error
Christopher Baines [Tue, 2 Aug 2011 09:29:11 +0000 (10:29 +0100)]
Inproved the SIGINT stop in the dpkgpm, not perfect yet but it should work when using Immediate-Configure-All.
Michael Vogt [Mon, 1 Aug 2011 13:32:16 +0000 (15:32 +0200)]
fix crash when P.Arch() was used but the cache got remapped
Michael Vogt [Mon, 1 Aug 2011 11:57:28 +0000 (13:57 +0200)]
merged from lp:~mvo/apt/mvo
Christopher Baines [Mon, 1 Aug 2011 11:27:10 +0000 (12:27 +0100)]
Only stop on SigInt if the system state is right (still needs more testing).
More inprovements to the package manager to prevent packages from being configured
twice.
Michael Vogt [Mon, 1 Aug 2011 10:57:04 +0000 (12:57 +0200)]
apt-pkg/pkgcachegen.cc: copy Arch to avoid segfault on cache remap
Michael Vogt [Mon, 1 Aug 2011 08:56:49 +0000 (10:56 +0200)]
apt-pkg/sourcelist.cc: GetListOfFilesInDir() fails if the dir does not exists, so test before using that
Christopher Baines [Sun, 31 Jul 2011 15:26:39 +0000 (16:26 +0100)]
Only allow interupts when using, Immediate-Configure-All.
TODO for dpkgpm:
Useful messages about the interupt, what was done to what packages and what was not done to what packages.
Only fail when the system is in a clean state, at the moment it will fail either a configure or install run.
Michael Vogt [Fri, 29 Jul 2011 16:40:34 +0000 (18:40 +0200)]
apt-pkg/contrib/sha2_internal.h: remove extern "C" to avoid symbol clash with libssl
Michael Vogt [Fri, 29 Jul 2011 14:33:20 +0000 (16:33 +0200)]
test/libapt/hashsums_test.cc: add a test for HashString and VerifyFile
Michael Vogt [Fri, 29 Jul 2011 13:45:06 +0000 (15:45 +0200)]
merged from debian-sid