IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
- implemented post-transaction filetriggers, loosely based on filetriggers.patch
from Mandriva Linux (see %_rpmlibdir/posttrans-filetriggers for details)
- implemented %_rpmlibdir/0ldconfig.filetrigger, so that packages with
shared libraries need not to invoke ldconfig(1) in they %%post-scriptlets
This should facilitate library upgrades, e.g. glibc-prthread.
Consider that we upgrade gcc-* and glibc-* packages; and glibc
has new subpackage glibc-pthread (with libpthread and librt
shared libraries).
Old order was:
D: ========== tsorting packages (order, #predecessors, #succesors, tree, depth)
D: 0 0 14 0 0 glibc-preinstall-2.8.90-alt3
D: 1 1 21 0 1 glibc-core-2.8.90-alt3
D: 2 1 9 0 2 cpp4.3-4.3.2-alt5
D: 3 1 5 0 2 libgcc4.3-4.3.2-alt5
D: 4 2 13 0 3 glibc-pthread-2.8.90-alt3
D: 5 2 20 0 4 glibc-core-debug-2.8.90-alt3
D: 6 2 17 0 4 glibc-gconv-modules-2.8.90-alt3
D: 7 2 16 0 4 glibc-locales-2.8.90-alt3
D: 8 2 15 0 4 glibc-nss-2.8.90-alt3
D: 9 2 12 0 4 glibc-timezones-2.8.90-alt3
D: 10 2 11 0 4 glibc-utils-2.8.90-alt3
D: 11 2 10 0 5 iconv-2.8.90-alt3
D: 12 8 22 0 6 glibc-2.8.90-alt3
D: 13 4 19 0 7 glibc-devel-2.8.90-alt3
D: 14 1 18 0 8 glibc-devel-static-2.8.90-alt3
D: 15 4 8 0 8 gcc4.3-4.3.2-alt5
D: 16 1 4 0 2 libgfortran4.3-4.3.2-alt5
D: 17 2 3 0 3 libgfortran4.3-devel-4.3.2-alt5
D: 18 3 6 0 4 gcc4.3-fortran-4.3.2-alt5
D: 19 2 2 0 3 libstdc++4.3-4.3.2-alt5
D: 20 2 1 0 4 libstdc++4.3-devel-4.3.2-alt5
D: ========== successors only (presentation order)
D: 21 3 7 0 5 gcc4.3-c++-4.3.2-alt5
Note that #succesors value is actually changed using the package index
in the input list of packages (on the command line): earlier packages
have higher values. This is called "successors from tsort are processed
in presentation order". E.g. when choosing to upgrade between cpp4.3,
libgcc4.3, and libgfortran4.3, cpp4.3 gets upgraded first. (The
collation is probably due to using shell glob on the command line.)
The problem is that, in cpp4.3 %post-script, some pthread-dependent
code might be called, and pthread shared library is simply mssing
at that point (after glibc-core upgrade and before glibc-pthread
install).
New order is:
D: ========== tsorting packages (order, #predecessors, #succesors, tree, depth)
D: 0 0 1 0 0 glibc-preinstall-2.8.90-alt3
D: 1 1 17 0 1 glibc-core-2.8.90-alt3
D: 2 1 3 0 2 libgcc4.3-4.3.2-alt5
D: 3 2 8 0 3 glibc-pthread-2.8.90-alt3
D: 4 2 2 0 4 glibc-gconv-modules-2.8.90-alt3
D: 5 2 2 0 4 glibc-nss-2.8.90-alt3
D: 6 2 1 0 5 iconv-2.8.90-alt3
D: 7 2 1 0 4 glibc-locales-2.8.90-alt3
D: 8 2 1 0 4 glibc-timezones-2.8.90-alt3
D: 9 2 1 0 4 glibc-utils-2.8.90-alt3
D: 10 8 1 0 5 glibc-2.8.90-alt3
D: 11 4 4 0 6 glibc-devel-2.8.90-alt3
D: 12 2 1 0 3 libstdc++4.3-4.3.2-alt5
D: 13 2 1 0 4 libstdc++4.3-devel-4.3.2-alt5
D: 14 1 1 0 2 cpp4.3-4.3.2-alt5
D: 15 4 2 0 3 gcc4.3-4.3.2-alt5
D: 16 1 1 0 2 libgfortran4.3-4.3.2-alt5
D: 17 2 1 0 3 libgfortran4.3-devel-4.3.2-alt5
D: ========== successors only (presentation order)
D: 18 2 0 0 4 glibc-core-debug-2.8.90-alt3
D: 19 1 0 0 7 glibc-devel-static-2.8.90-alt3
D: 20 3 0 0 4 gcc4.3-fortran-4.3.2-alt5
D: 21 3 0 0 4 gcc4.3-c++-4.3.2-alt5
Note that #succesors now indicates the number of immediate successors;
libgcc4.3 now has 3 immediate successors (glibc-pthread, gcc4.3, and
libstdc++4.3), while cpp4.3 and libgfortran4.3 have only one immediate
successor.
Also removed tools/rpmsort.c.
> The are various serial representations of a partially ordered set.
>
> The default is what I call "chainsaw", always emit the node that has
> the most children. The "chainsaw" heuristic tries to emit nodes that
> are depended upon as early as possible to localize interactions
> amongst packages, but really should be
> Always emit the node of the largest sub-tree.
> rather than the number of immediate children. I call this "buzzsaw".
>
> Anaconda has the constraint of changing cd's during install. So
> "presentation" ordering preserves the arrival ordering into a
> transaction set. Too bad that "presentation" ordering is often
> incorrect because of no loop analysis first.
https://lists.dulug.duke.edu/pipermail/rpm-devel/2005-June/000468.html
Both "Requires(post): foo" and "Requires: foo" should be kept.
The first dependency is only for %post-script: after the package
is installed, rpm allows to remove "foo" (though apt-get does not).
This is only a quick fix, perhaps the code should be rewritten.
For instance, if with have both "Requires(pre): foo" and
"Requires(post): foo", those can be folded into single
"Requires(pre,post): foo".
Both (F_OK|X_OK) cannot be tested at the same time, which means that
X_OK is X_OK for directories, too. And anyway there's a huge period
of time between X_OK test and exec. So I remove the test which cannot
be performed reliably. As a bonus, I also replace execv() with execvp().
Based on rpm5.org code. However, I do not introduce special variables
to distinguish between toplevel package and subpackages -- we can tell
the difference simply by checking if current "pkg" and "spec->packages"
point to the same record (which is toplevel package in the linked list).
This idiom is already used in e.g. parsePreamble().
This fixes invalid cpio entries for `/dir/.*' constructs.
Consider this:
%install
install -pD /dev/null %buildroot/root/.bashrc
%files
/root/.*
Old behaviour:
$ rpm -qlvp /home/at/RPM/RPMS/athlon/test-1.0-alt1.athlon.rpm
drwxr-xr-x 2 root root 0 Mar 30 01:47 /root/.
drwxr-xr-x 2 root root 0 Mar 30 01:47 /root/..
drwxr-xr-x 2 root root 0 Mar 30 01:47 /root/../root
-rwxr-xr-x 1 root root 0 Mar 29 16:51 /root/../root/.bashrc
-rwxr-xr-x 1 root root 0 Mar 29 16:51 /root/./.bashrc
-rwxr-xr-x 1 root root 0 Mar 29 16:51 /root/.bashrc
$
New behaviour:
$ LD_LIBRARY_PATH=... rpm -bi test.spec
...
Processing files: test-1.0-alt1
error: File path must be canonical: /root/.
error: File path must be canonical: /root/..
RPM build errors:
File path must be canonical: /root/.
File path must be canonical: /root/..
$
The problem here is that glob(3) `dir/.*' globs both '.' and '..'.
It is not clear whether or not glob(3) should glob '.' and '..'.
$ bash -c 'echo /var/empty/.*'
/var/empty/. /var/empty/..
$ ksh -c 'echo /var/empty/.*'
/var/empty/.*
$
