Re: [CVS] RPM: rpm/ CHANGES VENDOR rpm/lib/ depends.c
It mostly doesn't matter whether enum's and such are typedef'd or not.
__BUT__
rpmlog is based on syslog(3) and there's really no reason not
to continue exactly that (yes ancient and still widely used) API
in rpm itself.
From man 3 syslog, the 1st arg to syslog(3) is an int, with multiple
sub-fields
that are supposed to be OR'd together.
And only one of those bit-fields corresponds to rpmlogLvlType.
So technically URPMI, not rpmlog, in rpm is b0rken if it is using
only rpmlogLvlType, and the prototype change you made in
lib/depends.c is incorrect. (re-read the very 1st line please).
Where the issue __DOES__ get to be enormously painful is
when C++ linkage from, say, apt-rpm is attempted. C++ does
not have the same promotion rules as C.
And so what does (and will) happen, if adding private typedef's for
every steenkin'
integer in rpm is pursued, is that the typedef's will have to be
exposed in includes. ATM rpm is nowhere near the point where a stable
API/ABI can be maintained. Sure I'd like to do to have a stable API/
ABI, but every
rearrangement I make has a ripple effect, and noone is working
on the global API/ABI design issues but me.
What happens instead is a This used to work! bug report, and some
change is hacked in, and that's how rpm arrived at this state of
affairs in the
first place, organically incremental creeping crud.
73 de Jeff
On Nov 1, 2008, at 3:32 PM, Per Øyvind Karlsen wrote:
RPM Package Manager, CVS Repository
http://rpm5.org/cvs/
Server: rpm5.org Name: Per Øyvind Karlsen
Root: /v/rpm/cvs Email: [EMAIL PROTECTED]
Module: rpm Date: 01-Nov-2008 20:32:06
Branch: HEAD Handle: 2008110119320402
Modified files:
rpm CHANGES VENDOR
rpm/lib depends.c
Log:
Mandriva: make loglevel for dependency loop detection
overridable, debug-level
by default. (loop-detection-optional-loglevel)
Summary:
RevisionChanges Path
1.2639 +6 -4 rpm/CHANGES
2.62+9 -0 rpm/VENDOR
1.427 +4 -0 rpm/lib/depends.c
patch -p0 '@@ .'
Index: rpm/CHANGES
=
=
=
=
=
=
==
$ cvs diff -u -r1.2638 -r1.2639 CHANGES
--- rpm/CHANGES30 Oct 2008 16:04:19 - 1.2638
+++ rpm/CHANGES1 Nov 2008 19:32:04 - 1.2639
@@ -1,15 +1,17 @@
5.2a2 - 5.2a3:
+- proyvind: Mandriva: make loglevel for dependency loop
detection
+ overridable, debug-level by default. (loop-detection-optional-
loglevel)
- jbj: fix: __OpenBSD__ insque(3) needs rpmsq rock initialized.
- jbj: add #ifdef's needed to compile on __OpenBSD__.
- proyvind: Mandriva: make directory symlink dependencies
optional with
-%{_check_symlink_deps} %{_check_dirname_deps}. (optional-
dirname-and-symlink-deps)
+ %{_check_symlink_deps} %{_check_dirname_deps}. (optional-
dirname-and-symlink-deps)
- proyvind: Mandriva: prefer file extension for detecting file
compression.
-(extension-based-compression-detection)
+ (extension-based-compression-detection)
- proyvind: Mandriva: disable buildrequires checking for
building src.rpms.
-(no-deps-on-building-srpms)
+ (no-deps-on-building-srpms)
- proyvind: Mandriva: disable preservation of file ownership
for source rpms.
-(no-owner-group-on-srpm-install)
+ (no-owner-group-on-srpm-install)
- rpm.org: Add posix.mkstemp() to Lua posix lib.
- jbj: fix: apply _dependency_whiteout to erased relations too.
- jbj: lua: display rudimentary info about lua with --showrc.
@@ .
patch -p0 '@@ .'
Index: rpm/VENDOR
=
=
=
=
=
=
==
$ cvs diff -u -r2.61 -r2.62 VENDOR
--- rpm/VENDOR 28 Oct 2008 15:25:19 - 2.61
+++ rpm/VENDOR 1 Nov 2008 19:32:04 - 2.62
@@ -544,6 +544,15 @@
~always be an issue.
+ Change: loop-detection-optional-loglevel
+ Purpose:Report dependency loops at optional log level
+ Reason: Mandriva has many dependency loops by
intentetion,
+ especially due to it's library package
policies.
+ For now we want to be able to not display
these as errors,
+ but rather as debug-level messages by
default, or at any
+ other log level defined by
%_loop_detection_loglevel.
+
+
o Name:
Re: [CVS] RPM: rpm/ CHANGES VENDOR rpm/lib/ depends.c
2008/11/1 Jeff Johnson [EMAIL PROTECTED] It mostly doesn't matter whether enum's and such are typedef'd or not. __BUT__ rpmlog is based on syslog(3) and there's really no reason not to continue exactly that (yes ancient and still widely used) API in rpm itself. From man 3 syslog, the 1st arg to syslog(3) is an int, with multiple sub-fields that are supposed to be OR'd together. And only one of those bit-fields corresponds to rpmlogLvlType. So technically URPMI, not rpmlog, in rpm is b0rken if it is using only rpmlogLvlType, and the prototype change you made in lib/depends.c is incorrect. (re-read the very 1st line please). Where the issue __DOES__ get to be enormously painful is when C++ linkage from, say, apt-rpm is attempted. C++ does not have the same promotion rules as C. And so what does (and will) happen, if adding private typedef's for every steenkin' integer in rpm is pursued, is that the typedef's will have to be exposed in includes. ATM rpm is nowhere near the point where a stable API/ABI can be maintained. Sure I'd like to do to have a stable API/ABI, but every rearrangement I make has a ripple effect, and noone is working on the global API/ABI design issues but me. What happens instead is a This used to work! bug report, and some change is hacked in, and that's how rpm arrived at this state of affairs in the first place, organically incremental creeping crud. 73 de Jeff I guess you're commenting on the previous commit, and not this one? :) I changed the type since I figured that it made the context and valid values more clear, while also nicer for debugging (ie. gdb displaying the enumated value ie. name such as RPMLOG_WARN rather than just the integer) and otherwise as you point out, it mostly doesn't matter :) For URPMI usage of rpmlog, I doubt it using the typedef'd type, this was a purely URPMI non-related commit. ;)
Re: [CVS] RPM: rpm/ CHANGES VENDOR rpm/lib/ depends.c
On Nov 1, 2008, at 4:18 PM, Per Øyvind Karlsen wrote: I guess you're commenting on the previous commit, and not this one? :) Likely. I changed the type since I figured that it made the context and valid values more clear, while also nicer for debugging (ie. gdb displaying the enumated value ie. name such as RPMLOG_WARN rather than just the integer) and otherwise as you point out, it mostly doesn't matter :) And I'm not disagreeing re context and/or valid values and/or (I'm glad you've noticed) debugging is way way easier with typedef'd enums. But note rpmMireMode, what could be easier or more innocent than an enum, right? Wrong. Once the enum gets used in a prototype, then the typedef needs to be included for rpmdb.h. Which drags mire.h into rpm's API for no particularly important reason. For many reasons, I'd like mire.h to be stand-alone with the typedef''d enum within mire.h. And for other, conflicting reasons, I'd like not to expose an API that may (or may not) include functionality like PCRE, because PCRE cannot be used unless --with-pcre is included. I don't pretend to have answers here. For URPMI usage of rpmlog, I doubt it using the typedef'd type, this was a purely URPMI non-related commit. ;) Good. 73 de Jeff__ RPM Package Managerhttp://rpm5.org Developer Communication Listrpm-devel@rpm5.org
Re: [CVS] RPM: rpm/ CHANGES configure.ac
What is wrong with CPPFLAGS=-DRPM_VENDOR_MANDRIVA Its insane to have Yet More Ways to configure stuff imho. 73 de Jeff On Nov 1, 2008, at 6:41 PM, Per Øyvind Karlsen wrote: RPM Package Manager, CVS Repository http://rpm5.org/cvs/ Server: rpm5.org Name: Per Øyvind Karlsen Root: /v/rpm/cvs Email: [EMAIL PROTECTED] Module: rpm Date: 01-Nov-2008 23:41:44 Branch: HEAD Handle: 2008110122414300 Modified files: rpm CHANGES configure.ac Log: add --with-vendor configure option to define RPM_VENDOR_*. Summary: RevisionChanges Path 1.2640 +1 -0 rpm/CHANGES 2.319 +28 -0 rpm/configure.ac patch -p0 '@@ .' Index: rpm/CHANGES = = = = = = == $ cvs diff -u -r1.2639 -r1.2640 CHANGES --- rpm/CHANGES1 Nov 2008 19:32:04 - 1.2639 +++ rpm/CHANGES1 Nov 2008 22:41:43 - 1.2640 @@ -1,5 +1,6 @@ 5.2a2 - 5.2a3: +- proyvind: add --with-vendor configure option to define RPM_VENDOR_*. - proyvind: Mandriva: make loglevel for dependency loop detection overridable, debug-level by default. (loop-detection-optional- loglevel) - jbj: fix: __OpenBSD__ insque(3) needs rpmsq rock initialized. @@ . patch -p0 '@@ .' Index: rpm/configure.ac = = = = = = == $ cvs diff -u -r2.318 -r2.319 configure.ac --- rpm/configure.ac 19 Oct 2008 07:46:48 - 2.318 +++ rpm/configure.ac 1 Nov 2008 22:41:43 - 2.319 @@ -1636,6 +1636,34 @@ AC_SUBST(RPMLUAFILES) AC_MSG_RESULT([$RPMLUAFILES]) +dnl # determine whether to build using a specific vendor's changes or not +AC_MSG_CHECKING([if building for a specfic vendor]) + +AC_ARG_WITH( +[vendor], +AS_HELP_STRING([--with-vendor=VENDOR], [build with a supported vendor's specific set of changes: openpkg, rpm4darwin, fedora, mandriva or windriver]), +if test .$withval != .no; then + if test $withval == openpkg; then + AC_DEFINE([RPM_VENDOR_OPENPKG], 1, Vendor is OpenPKG) + fi + if test $withval == rpm4darwin; then + AC_DEFINE([RPM_VENDOR_RPM4DARWIN], 1, Vendor is RPM for Darwin) + fi + if test $withval == fedora; then + AC_DEFINE([RPM_VENDOR_FEDORA], 1, Vendor is Fedora) + fi + if test $withval == mandriva; then + AC_DEFINE([RPM_VENDOR_MANDRIVA], 1, Vendor is Mandriva) + fi + if test $withval == windriver; then + AC_DEFINE([RPM_VENDOR_WINDRIVER], 1, Vendor is Windriver) + fi + AC_MSG_RESULT([$withval]) +else + AC_MSG_RESULT(no) +fi +) + testdir=`pwd`/tests AC_SUBST(testdir) @@ . __ RPM Package Managerhttp://rpm5.org CVS Sources Repository[EMAIL PROTECTED] __ RPM Package Managerhttp://rpm5.org Developer Communication Listrpm-devel@rpm5.org
Re: [CVS] RPM: rpm/ CHANGES configure.ac
2008/11/1 Jeff Johnson [EMAIL PROTECTED] What is wrong with CPPFLAGS=-DRPM_VENDOR_MANDRIVA Its insane to have Yet More Ways to configure stuff imho. 73 de Jeff It's hidden from regular user who'd like to build rpm himself for that specific vendor. Maybe not such an important thing for rpm devels, but I know that if I were to build rpm5 myself for ie. Mandriva without being involved in rpm development nor reading through source just for fun, I'd like to easily find and use the suitable option to pick what to be considered the most proper adaptions using './configure --help'. Moreover, personally, I just find it nicer and more intuitive than passing either CPPFLAGS, or adding it to config.h.. ;)
Re: [CVS] RPM: rpm/ CHANGES configure.ac
On Nov 1, 2008, at 7:19 PM, Per Øyvind Karlsen wrote: 2008/11/1 Jeff Johnson [EMAIL PROTECTED] What is wrong with CPPFLAGS=-DRPM_VENDOR_MANDRIVA Its insane to have Yet More Ways to configure stuff imho. 73 de Jeff It's hidden from regular user who'd like to build rpm himself for that specific vendor. Hidden? Adding another option is no substitute for identifying exactly the functionality desired in Mandriva. If you want better doco, then write better doco on how to use rpm5.org on Mandriva. Adding --with-vendor does nothing except what CPPFLAGS=-DRPM_VENDOR_MANDRIVA already does. I don't see any additional benefit from adding an enumeration for vendors that are, in fact, hardly likely to adopt rpm5.org any time soon. Not that Mandriva uses rpm5.org, nor is Mandriva expected to change from rpm.org. So what purpose is served? If --with-vendor=mandriva did anything useful, like identifying specific components that should be included when building, I might be saying something else. But there is already devtool, which can add an entire collection of configure options when invoking rpm5.org ./configure. Try adding a %mandriva stanza to devtool.conf. Maybe not such an important thing for rpm devels, but I know that if I were to build rpm5 myself for ie. Mandriva without being involved in rpm development nor reading through source just for fun, I'd like to easily find and use the suitable option to pick what to be considered the most proper adaptions using './configure -- help'. I couldn't care less about yjr AutoFu. But adding --with-vendor=fedora is hardly the same thing as claiming Fedora support. Fedora has most definitely said that rpm5.org is not permitted in the Fedora distro without being crippled. I see no reason to pretend support --with-vendor=fedora for a distro that wants to cripple rpm5.org functionality. And I'm quite sure that rpm5.org would not be permitted in Fedora no matter what. The criteria would just change to Now that we've crippled rpm5.org so that its useless, do we really need to include in Fedora? Moreover, personally, I just find it nicer and more intuitive than passing either CPPFLAGS, or adding it to config.h.. ;) So have at. But I don't see how adding Yet More AutoFu changes much of anything at all. 73 de Jeff
Re: [CVS] RPM: rpm/ CHANGES Makefile.am rpm/tests/ Makefile.am
Adding test *.src.rpm packages does add a certain amount of bloat to the rpm distribution tarball. An alternative approach would require a reference URI, and tie make check to wget and network access. Note that make check already fetches popt-1.14-1.src.rpm, so it would not be new behavior. OTOH, there are those who likely do not wish a make check to do network accesses. I dunno where the right balance point is between make check and bloat yet. I'm open to suggestions. Meanwhile, make check has certainly not been functional outside of rpm5.org CVS yet. 73 de Jeff On Nov 1, 2008, at 10:28 PM, Jeff Johnson wrote: RPM Package Manager, CVS Repository http://rpm5.org/cvs/ Server: rpm5.org Name: Jeff Johnson Root: /v/rpm/cvs Email: [EMAIL PROTECTED] Module: rpm Date: 02-Nov-2008 03:28:06 Branch: HEAD Handle: 2008110202280401 Modified files: rpm CHANGES Makefile.am rpm/tests Makefile.am Log: - include tests/*.src.rpm, exclude tests/ref/CVS/*, with make dist. Summary: RevisionChanges Path 1.2643 +1 -0 rpm/CHANGES 2.223 +11 -1 rpm/Makefile.am 1.34+3 -1 rpm/tests/Makefile.am patch -p0 '@@ .' Index: rpm/CHANGES = = = = = = == $ cvs diff -u -r1.2642 -r1.2643 CHANGES --- rpm/CHANGES2 Nov 2008 00:55:29 - 1.2642 +++ rpm/CHANGES2 Nov 2008 02:28:04 - 1.2643 @@ -1,5 +1,6 @@ 5.2a2 - 5.2a3: +- jbj: include tests/*.src.rpm, exclude tests/ref/CVS/*, with make dist. - proyvind: lua: explicitly link against libexpat as assumption on implicit linking cannot be guaranteed. - proyvind: add --with-extra-path-macros configure option that adds @@ . patch -p0 '@@ .' Index: rpm/Makefile.am = = = = = = == $ cvs diff -u -r2.222 -r2.223 Makefile.am --- rpm/Makefile.am21 Aug 2008 13:15:29 - 2.222 +++ rpm/Makefile.am2 Nov 2008 02:28:05 - 2.223 @@ -32,7 +32,17 @@ python/mpw/Makefile.am python/mpw/test/* python/test/resources/* python/test/dscmp \ python/test/psTest.py python/test/testhdr python/test/testit python/test/testmi \ python/test/testrollback.py python/test/dsMerge.py \ - python/rpm/__init__.py python/[a-z]* python/Makefile.in python/ Makefile.am + python/rpm/__init__.py python/[a-z]* python/Makefile.in python/ Makefile.am \ + tests/devtool-sanity-1.0-1.src.rpm \ + tests/edos-test-1-0.src.rpm \ + tests/probes-test-1-0.src.rpm \ + tests/triggers-D-1.0-1.src.rpm \ + tests/triggers-DP-1.0-1.src.rpm \ + tests/triggers-F-1.0-1.src.rpm \ + tests/triggers-FP-1.0-1.src.rpm \ + tests/triggers-N-1.0-1.src.rpm \ + tests/triggers-NA-1.0-1.src.rpm \ + tests/triggers-P-1.0-1.src.rpm SUBDIRS = if USE_NLS @@ . patch -p0 '@@ .' Index: rpm/tests/Makefile.am = = = = = = == $ cvs diff -u -r1.33 -r1.34 Makefile.am --- rpm/tests/Makefile.am 15 Oct 2008 18:15:31 - 1.33 +++ rpm/tests/Makefile.am 2 Nov 2008 02:28:05 - 1.34 @@ -2,7 +2,9 @@ AUTOMAKE_OPTIONS = 1.4 foreign -EXTRA_DIST = ref/* ref/.alldigests +# Note: *.src.rpm's cannot be added here because of suffix rules. +EXTRA_DIST = ref/[^C]* ref/.alldigests + SUBDIRS = macros = $(abs_top_builddir)/macros:$(testdir)/macros @@ . __ RPM Package Managerhttp://rpm5.org CVS Sources Repository[EMAIL PROTECTED] __ RPM Package Managerhttp://rpm5.org Developer Communication Listrpm-devel@rpm5.org
