The funcs xbps_pkg_name() and xbps_pkgpattern_name() were
using malloc(3) to return the result, until now.
They now have been changed to not allocate the result
via malloc, the caller is responsible to provide a buffer
at least of XBPS_NAME_SIZE (64).
If for whatever reason the pkgname can't be guessed,
returns false. This should avoid lots of small allocs
around libxbps.
New functions have the following prototype:
bool xbps_pkg_name(char *dst, size_t len, const char *pkg)
bool xbps_pkgpattern_name(char *dst, size_t len, const char *pkg)
as suggested by @duncaen.
This fixes the keep_provider_on_update test case, submitted
by @st3r4g via #219.
While checking for obsoletes make sure current provider
also matches the current target pkg.
Fixes#219
While looking for dependencies, we need to check
if xbps_rpool_get_pkg() returned a suitable match;
and then validate its result.
This fixes the update_and_install test case that
was reverted via #218.
This reverts a change that solved another issue and marks the test case
for the other issue as expected failure.
The other issue is not as important as this, as it blocks updating a lot
of systems.
While checking for orphans in automatic mode (xbps-query -O,
xbps-remove -o) there's no need to generate a fulldeptree...
instead we need to iterate against pkgdb until no more orphans
are found.
See https://github.com/void-linux/xbps/issues/156#issuecomment-578473222
Also add some debugging that helped me to catch the issue
easily.
Closes#156
Added support to download all dependencies even
if the euid does not have write perms to rootdir.
In this mode we only care if cachedir is writable,
rootdir access is not necessary.
This is really useful to download all binary packages
required by any number of packages as any regular
user to later perform off-line installations, i.e:
```
$ xbps-install -c $PWD/cachedir -yD xbps
...
$ tree cachedir
cachedir/
├── acl-2.2.53_1.x86_64-musl.xbps
├── acl-2.2.53_1.x86_64-musl.xbps.sig
├── attr-2.4.48_1.x86_64-musl.xbps
├── attr-2.4.48_1.x86_64-musl.xbps.sig
├── bzip2-1.0.8_1.x86_64-musl.xbps
├── bzip2-1.0.8_1.x86_64-musl.xbps.sig
├── ca-certificates-20190110_1.noarch.xbps
├── ca-certificates-20190110_1.noarch.xbps.sig
├── libarchive-3.4.1_1.x86_64-musl.xbps
├── libarchive-3.4.1_1.x86_64-musl.xbps.sig
├── libcrypto45-3.0.2_2.x86_64-musl.xbps
├── libcrypto45-3.0.2_2.x86_64-musl.xbps.sig
├── liblz4-1.9.2_1.x86_64-musl.xbps
├── liblz4-1.9.2_1.x86_64-musl.xbps.sig
├── liblzma-5.2.4_2.x86_64-musl.xbps
├── liblzma-5.2.4_2.x86_64-musl.xbps.sig
├── libressl-3.0.2_2.x86_64-musl.xbps
├── libressl-3.0.2_2.x86_64-musl.xbps.sig
├── libssl47-3.0.2_2.x86_64-musl.xbps
├── libssl47-3.0.2_2.x86_64-musl.xbps.sig
├── libtls19-3.0.2_2.x86_64-musl.xbps
├── libtls19-3.0.2_2.x86_64-musl.xbps.sig
├── libxbps-0.57.1_8.x86_64-musl.xbps
├── libxbps-0.57.1_8.x86_64-musl.xbps.sig
├── libzstd-1.4.4_1.x86_64-musl.xbps
├── libzstd-1.4.4_1.x86_64-musl.xbps.sig
├── musl-1.1.24_1.x86_64-musl.xbps
├── musl-1.1.24_1.x86_64-musl.xbps.sig
├── run-parts-4.9.1_1.x86_64-musl.xbps
├── run-parts-4.9.1_1.x86_64-musl.xbps.sig
├── xbps-0.57.1_8.x86_64-musl.xbps
├── xbps-0.57.1_8.x86_64-musl.xbps.sig
├── xbps-triggers-0.113_3.noarch.xbps
├── xbps-triggers-0.113_3.noarch.xbps.sig
├── zlib-1.2.11_3.x86_64-musl.xbps
└── zlib-1.2.11_3.x86_64-musl.xbps.sig
0 directories, 36 files
$
```
Inpired by #213Closes#213
When there's a new xbps update, xbps-install(1) will now return
EBUSY (16) and a message (if dry-run disabled) explaining
how to proceed.
If there's an update and transaction does not contain xbps, it will
error out unless the 'xbps' pkg is the only target pkg, i.e:
# xbps-install -Su
# echo $?
16
To update xbps, the only way to proceed is to explicitly declare
it as an update, i.e:
# xbps-install -u xbps
The dry-run mode will still show there's an xbps update.
Modified the existing test cases to satisfy the new behaviour.
Closes#166Closes#142
If all packages in transaction are on hold, there's no
need to check for conflicts, shlibs, etc.
This makes `xbps-install -un` work faster on my system:
~0.450ms -> ~0.250ms
There's still room for more improvements :-)
See the new test case for details, but this simulates
the recent glvnd switch, with mesa, nvidia and libglvnd.
Thanks @st3r4g for the fix! this indeed fixes the new test case.
Close#186
We do not want to continue processing more threads
if pthread_create(3) fails, rather return an error.
This is for #182 but not yet fixed, there might be a
memleak somewhere.
xbps-rindex(1) has a -r option that sets the remove mode,
due to the changes added to sign repodata we need to access
to the correct directory where repository public keys are
stored. This makes the code use `$PWD/keys` before falling
back to `metadir`.
Fixes the test suite to run with unprivileged users (non root).
In the edge case when an updated package has different (or no)
alternatives groups, make sure to prune those that are in pkgdb
but not in the newly installed package.
A potentially common case of this is when a package that formerly
had alternatives gets removed and a transitional metapackage
takes its place (which has no alternatives).
When the new package has no dependencies, oldest next possible
alternatives group will be used. This is because that indicates
a removed package. When there are dependencies, the newest one
will be used; as this indicates a transitional package.
This only changes the debug messages, because modified files and conf
files are handled the same on removal and unless a conf file changes
the file type it is never deleted.
`xbps_transaction_revdeps` will now try to add updates for packages to the
transaction if the dependency is not satisified anymore due to a package
install/update.
`xbps_transaction_prepare` will now check the return value of
`xbps_transaction_revdeps` and repeat the dependency resolution,
until `xbps_transaction_revdeps` returns 0, which means that it didn't
add any new packages to the transaction.
fetch/http.c:1475: warning: function might return no value: 'parse_index'
The code uses abort() before returning anything; return a proper
return value instead.
This reverts commit 08aa44602f.
The element still has to be processed and ignored by one of the
threads, it is not the last entry this will make the actually last item
be ignored.
the first thread to finish will start again from 0 (`reserved`) until
0+slicecount, next thread will then start from the end of the previous
thread which is already wrong, resulting in processing the first
slicecount*maxthreads entries twice.
The first slicecount*maxthreads entries are accounted by the thread
creation loop and `reserved` has to start at the first unaccounted index.
Added `-D, --download-only` flag to allow downloading
packages to the cache without attempting to install them.
Contributed by Toyam Cox via #9
Adapted to master by yours truly.
Close#9
If set it will continue with the transaction and will
just print what are the conflicting files without returning
EEXIST.
This is a temporary solution for void where there are still
some packages with conflicting files (qt5-host-tools vs qt5-tools-devel).
With input by @duncaen
On some systems, something like this happens:
package_unpack.c:375:11: error: 'buf' may be used uninitialized in this function
This is not actually a bug as logically `buf` is always initialized
in that place, but gcc doesn't like it anyway.
if xbps_repo_open() fails then the repo is removed from array
xbps_repo_remove() but as we've removed the entry from it,
the index won't be valid any more, resulting in skipped entries.
If a entry is removed restart again at the correct index.
Close#127
Found by clang analyzer.
Bug Summary
File: lib/portableproplib/rb.c
Warning: line 363, column 19
Value stored to 'grandpa' during its initialization is never read
It was made many years ago to update the pkgdb format, but
it's innecessary to build it by default because is dead code.
Anyway keep it just in case pkgdb format changes.
Initialize rv to avoid a logic error.
Found by clang-analyzer.
Bug Summary
File: lib/initend.c
Warning: line 101, column 10
The left operand of '<' is a garbage value
Reported by clang-analyzer.
After reading the linux man-pages vfork(2) manpage I came to the
conclusion that it's better to change the behaviour of the
original source (NetBSD) and change this to use fork(2).
Always check if there's a new xbps package version available
while *installing* or *updating* in any form.
This fixes the following scenario:
- xbps-0.53_10 is currently installed
- xbps-0.54_1 is available in repo
- xbps-install --update netbsd-wtf || xbps-install new-pkg || xbps-install --update
As expected any of the following scenarios in last cmd will use *this*
transaction to autoupdate xbps and its reverse dependencies.
Another transaction will be necessary to install or update the other unrelated
packages.
Added a new test case to verify this case and improve the other test
cases with more extensive checks.
xbps_get_pkg_fulldeptree() now returns NULL and sets errno to ENODEV
when there are missing dependencies, rather than assert()ing.
Added another test case to check returned error codes.
Signed-off-by: Juan RP <xtraeme@gmail.com>
As shown in https://github.com/void-linux/xbps/issues/110
/var might be rw, while / might be ro.
Check this via xbps_pkgdb_lock() for both things.
Close#110
Signed-off-by: Juan RP <xtraeme@gmail.com>
The current code was failing because while checking for updates
on its reverse dependencies, up-to-date versions were treated
as an error.
Added a new test case to verify that it works as expected.
Signed-off-by: Juan RP <xtraeme@gmail.com>
