Bug#1071429: dpkg: `dpkg-db-backup.timer` starts `dpkg-db-backup.service` in hot path of boot

Hi! On Sun, 2024-05-19 at 07:09:24 +0200, Paul Menzel wrote: > Package: dpkg > Version: 1.22.6 > Severity: normal > Trying to decrease the boot time of a desktop system (pressing the power > button to GDM login screen), I noticed `dpkg-db-backup.service` in the > hotpath: […] > It’d be great

[dpkg] 06/08: libdpkg: Compute string slice for varbuf_has_suffix() check on good values

Author: Guillem Jover AuthorDate: Sun May 19 14:12:58 2024 +0200 libdpkg: Compute string slice for varbuf_has_suffix() check on good values We should not compute the slice with «used» members values that we know will produce a wrong pointer, even if we do not end up using

[dpkg] 07/08: libdpkg: Make prefix and suffix varbuf checks cope with unallocated varbufs

Author: Guillem Jover AuthorDate: Sun May 19 11:21:10 2024 +0200 libdpkg: Make prefix and suffix varbuf checks cope with unallocated varbufs If the varbufs have not been allocated, we should not be passing them to functions that expect their arguments not to be NULL. Changelog

[dpkg] 05/08: man: Update verify format example to also include M

Author: Guillem Jover AuthorDate: Wed May 15 02:30:35 2024 +0200 man: Update verify format example to also include M We implemented partial support for the M flag in the output, so mention it in the example format to make it clear it is supported. --- man/dpkg.pod | 2 +- 1 file

[dpkg] 08/08: libdpkg: Make varbuf_detach() always return a string

-> main) Author: Guillem Jover AuthorDate: Fri May 17 23:35:26 2024 +0200 libdpkg: Make varbuf_detach() always return a string When we have added no content into a varbuf, we avoid allocating memory, but when calling varbuf_detach() the caller expects a valid string no mat

[dpkg] 01/08: test: Refactor real and virtual package setup

Author: Guillem Jover AuthorDate: Sun May 12 19:27:05 2024 +0200 test: Refactor real and virtual package setup Use a list of real and virtual packages to use, and feed them into the known facts object in a loop. Based-on-patch-by: Johannes Schauer Marin Rodrigues --- scripts

[dpkg] 04/08: dpkg: Do not run hooks or loggers with --dry-run or unprivileged

Author: Guillem Jover AuthorDate: Wed May 15 01:55:57 2024 +0200 dpkg: Do not run hooks or loggers with --dry-run or unprivileged If we have been told to run in dry mode, we should not be executing the invocation hooks, as those might have side effects. We should not run either

[dpkg] 03/08: debian: Suppress s-s-d compat symlink if /sbin is missing

by the bootstrapping tool, but that would required encoding this ordering somehow. Closes: #1071078 Signed-off-by: Guillem Jover --- debian/dpkg.postinst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/debian/dpkg.postinst b/debian/dpkg.postinst index 113c8d53a..8d9f6067f 100644

[dpkg] 02/08: dpkg-genbuildinfo: Parse Provides as virtual packages

Author: Guillem Jover AuthorDate: Sat May 11 21:12:33 2024 +0200 dpkg-genbuildinfo: Parse Provides as virtual packages We did not specify that the dependencies to be parsed were virtual packages. --- scripts/dpkg-genbuildinfo.pl | 6 +- 1 file changed, 5 insertions(+), 1 deletion

[dpkg] branch main updated (b33121e1f -> 17dd89881)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from b33121e1f po: Update German man pages translation new b126d9726 test: Refactor real and virtual package setup new 98e16e2eb dpkg-genbuildinfo: Parse

Bug#1071124: dpkg: post-invoke hook does not run after failure due to non-superuser (but pre-invoke hook does)

Hi! On Tue, 2024-05-14 at 18:19:18 +, Ben Ferreri wrote: > Package: dpkg > Version: 1.22.6 > Severity: normal > X-Debbugs-Cc: bferr...@intwineconnect.com > When attempting to install a package as a non-superuser, the configured > pre-invoke hook runs, but the post-invoke hook does not. I

Bug#1071078: Skip creation of start-stop-daemon compat symlink if /sbin does not exist

Hi! On Mon, 2024-05-13 at 22:36:12 +0200, Johannes Schauer Marin Rodrigues wrote: > Package: dpkg > Version: 1.21.22 > Severity: wishlist > Tags: patch > dpkg postinst fails to create the start-stop-daemon compatibility > symlink if /usr does not exist: > > ln: failed to create symbolic

[dpkg] 02/04: man: Document weak checksum algorithms

Author: Guillem Jover AuthorDate: Tue Jun 6 23:56:25 2023 +0200 man: Document weak checksum algorithms Add an attribute describing whether the fields are weak or strong, and a note explicitly explaining their secure verification status. --- man/deb-buildinfo.pod | 9 ++--- man

[dpkg] 03/04: libdpkg: Refactor lax problem reporting into parse_lax_problem() function

Author: Guillem Jover AuthorDate: Fri May 10 00:18:47 2024 +0200 libdpkg: Refactor lax problem reporting into parse_lax_problem() function This function checks whether need to report a warning or an error depending on whether we are currently in one of the specific lax modes

[dpkg] 01/04: man: Document missing Packages front-end fields in dpkg-query(1)

Author: Guillem Jover AuthorDate: Tue Jun 6 23:54:05 2023 +0200 man: Document missing Packages front-end fields in dpkg-query(1) These are part of the Packages files in repositories, which can end up in the available database shown by dpkg-query(1). --- man/dpkg-query.pod | 8

[dpkg] branch main updated (77f1ce33a -> b33417497)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from 77f1ce33a po: Update Dutch man pages translations new e5fb453b3 man: Document missing Packages front-end fields in dpkg-query(1) new 4ed783dc7 man:

[dpkg] branch main updated: po: Update Dutch man pages translations

ses: #1070144 Signed-off-by: Guillem Jover --- man/po/nl.po | 401 +-- 1 file changed, 250 insertions(+), 151 deletions(-) diff --git a/man/po/nl.po b/man/po/nl.po index 1ff3a9348..9b8c4d556 100644 --- a/man/po/nl.po +++ b/man/po/nl

Re: Architecture variants for Debian / Ubuntu

On Sat, 2024-05-04 at 01:07:59 +0200, Matthias Klose wrote: > On 03.05.24 11:27, Julian Andres Klode wrote: > > On Wed, Sep 06, 2023 at 11:27:02AM +0200, Guillem Jover wrote: > > > Some of the above problems could perhaps be avoided if we introduced > > > a concept o

[dpkg] 07/12: test: Test variable override

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:16:20 2024 +0200 test: Test variable override Signed-off-by: Guillem Jover --- scripts/t/mk/architecture.mk | 3 +++ 1 file changed, 3 insertions(+) diff --git a/scripts/t/mk/architecture.mk b/scripts/t/mk/architecture.mk index

[dpkg] 06/12: test: Test exported variables in addition to Make variables

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:14:27 2024 +0200 test: Test exported variables in addition to Make variables [guil...@debian.org: - Fix comments format. ] Signed-off-by: Guillem Jover --- scripts/t/mk/architecture.mk | 3 +++ scripts/t/mk/buildflags.mk

[dpkg] 05/12: test: Replace double quotes with single quote in shell recipes

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:27:16 2024 +0200 test: Replace double quotes with single quote in shell recipes This escapes more characters for the shell, and makes the intent of the dollar character more explicit. Signed-off-by: Guillem Jover

[dpkg] 11/12: scripts/mk: Reduce the number of subprocesses

mechanism for consistency. [guil...@debian.org: - Rename dpkg_lazy to dpkg_lazy_eval. - Reword commit message. - Fix indentation. ] Closes: #872381 Signed-off-by: Guillem Jover --- scripts/mk/architecture.mk | 15 ++- scripts/mk/buildapi.mk | 6

[dpkg] 12/12: libdpkg: Mark file_slurp_fd() NUL-terminate the varbuf

-> main) Author: Guillem Jover AuthorDate: Sat May 4 02:27:44 2024 +0200 libdpkg: Mark file_slurp_fd() NUL-terminate the varbuf While this is not a concern with the current users of these functions, it's still better to NUL-terminate the resulting varbuf, in case a cal

[dpkg] 10/12: test: Test override of a build tool

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:23:19 2024 +0200 test: Test override of a build tool Signed-off-by: Guillem Jover --- scripts/t/mk/buildtools.mk | 4 1 file changed, 4 insertions(+) diff --git a/scripts/t/mk/buildtools.mk b/scripts/t/mk/buildtools.mk index

[dpkg] 08/12: test: Test DEB_CXXFLAGS_MAINT_SET

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:20:42 2024 +0200 test: Test DEB_CXXFLAGS_MAINT_SET Signed-off-by: Guillem Jover --- scripts/t/mk/buildflags.mk | 6 ++ 1 file changed, 6 insertions(+) diff --git a/scripts/t/mk/buildflags.mk b/scripts/t/mk/buildflags.mk index

[dpkg] 09/12: test: Add missing test for CPP build tool

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:21:58 2024 +0200 test: Add missing test for CPP build tool Signed-off-by: Guillem Jover --- scripts/t/mk/buildtools.mk | 1 + 1 file changed, 1 insertion(+) diff --git a/scripts/t/mk/buildtools.mk b/scripts/t/mk/buildtools.mk index

[dpkg] 01/12: build: Rework subst handling for built or installed artifacts

Author: Guillem Jover AuthorDate: Thu Apr 11 04:32:39 2024 +0200 build: Rework subst handling for built or installed artifacts Switch from sed to perl so that we can perform in-place substitutions in a portable way. Make the variable naming usage uniform. This reduces

[dpkg] 04/12: test: Use loops instead of repetitions in mk fragment tests

Author: Nicolas Boulenguez AuthorDate: Wed Apr 10 00:09:43 2024 +0200 test: Use loops instead of repetitions in mk fragment tests [guil...@debian.org: - Wrap line at foreach. ] Signed-off-by: Guillem Jover --- scripts/t/mk/architecture.mk | 54

[dpkg] 03/12: test: Simplify buildflags.mk test of _MAINT_APPEND when TEST_ is empty

do not need to care whether the TEST_ variables were previously set or not. Signed-off-by: Guillem Jover --- scripts/t/mk/buildflags.mk | 7 +-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/scripts/t/mk/buildflags.mk b/scripts/t/mk/buildflags.mk index 94d85a7e0

[dpkg] branch main updated (84d82aaaf -> 9dff67290)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from 84d82aaaf po: Update German programs translation new f1175056f build: Rework subst handling for built or installed artifacts new ce5c5af16 test: Unset

[dpkg] 02/12: test: Unset DEB_BUILD_MAINT_OPTIONS in build flags tests

Author: Guillem Jover AuthorDate: Mon May 6 00:42:46 2024 +0200 test: Unset DEB_BUILD_MAINT_OPTIONS in build flags tests If we set this environment variable with optimize=+lto, then the test fails. We need to unset it, like we do in other test files, to be resilient against users

[dpkg] branch main updated: Revert "test: Pass -T+1 to xz to workaround spurious warning with xz 5.6.0"

by this push: new 24a7aee46 Revert "test: Pass -T+1 to xz to workaround spurious warning with xz 5.6.0" 24a7aee46 is described below commit 24a7aee4631c8fb6f6482297810937f93bcc85da (HEAD -> main) Author: Guillem Jover AuthorDate: Mon Apr 29 03:47:24 2024 +0200 Revert &qu

[dpkg] branch main updated: test: Initialize varbuf before use

by this push: new 0f4000a29 test: Initialize varbuf before use 0f4000a29 is described below commit 0f4000a29e2c8fdbe007a645028c10a04bdebbcc (HEAD -> main) Author: Guillem Jover AuthorDate: Mon Apr 29 03:30:45 2024 +0200 test: Initialize varbuf before use This was already fi

[dpkg] branch main updated: libdpkg: Add missing header includes

by this push: new 489c0f4b2 libdpkg: Add missing header includes 489c0f4b2 is described below commit 489c0f4b2fcee6169640328fa8188174f2e4888e (HEAD -> main) Author: Guillem Jover AuthorDate: Mon Apr 29 03:17:52 2024 +0200 libdpkg: Add missing header includes Fixes: com

[dpkg] 15/17: src: Fix timestamp parse error reporting

Author: Guillem Jover AuthorDate: Thu Apr 25 23:02:31 2024 +0200 src: Fix timestamp parse error reporting We should distinguish between errors causing errno to be set, and errors from partial conversions or no conversions at all. Otherwise we might get an error message that states

[dpkg] 09/17: man: Document DPKG_COLORS and DPKG_NLS for all perl scripts honoring them

Author: Guillem Jover AuthorDate: Sat Apr 13 04:42:57 2024 +0200 man: Document DPKG_COLORS and DPKG_NLS for all perl scripts honoring them The dpkg-buildapi and dpkg-buildtree scripts honor these environment variables by way of the Dpkg::ErrorHandling and Dpkg::Gettext modules

[dpkg] 07/17: build: Pass abs_srcdir and abs_builddir to the TAP driver

Author: Guillem Jover AuthorDate: Sat Apr 20 19:28:54 2024 +0200 build: Pass abs_srcdir and abs_builddir to the TAP driver We will need at least the abs_builddir to add tests for getcwd(), so we add both out of completeness. --- build-aux/tap.am | 2 ++ 1 file changed, 2 insertions

[dpkg] 13/17: dpkg-realpath: Rewrite in C

Author: Guillem Jover AuthorDate: Sat Apr 13 22:51:39 2024 +0200 dpkg-realpath: Rewrite in C This should make the code more robust against system issues on missing shell interpreters or realpath and readlink commands. --- po/POTFILES.in | 2 + src/Makefile.am | 7

[dpkg] 17/17: po: Update Swedish translations

-> main) Author: Peter Krefting AuthorDate: Sun Apr 28 15:43:47 2024 +0100 po: Update Swedish translations Closes: #1070011 Signed-off-by: Peter Krefting Signed-off-by: Guillem Jover --- man/po/sv.po | 324 ---

[dpkg] 05/17: libdpkg: Always NUL terminate varbufs

Author: Guillem Jover AuthorDate: Wed Jul 12 04:41:33 2023 +0200 libdpkg: Always NUL terminate varbufs This guarantees the varbufs returned by functions are always printable and prepared to be handled by C string functions. --- lib/dpkg/t/t-varbuf.c | 14 -- lib/dpkg

[dpkg] 16/17: src: Check whether SOURCE_DATE_EPOCH is set before parsing it

Author: Guillem Jover AuthorDate: Thu Apr 25 22:44:19 2024 +0200 src: Check whether SOURCE_DATE_EPOCH is set before parsing it The dpkg-deb and dpkg-split program try to parse this environment variable to use it for their timestamps inside files to generate reproducible artifacts

[dpkg] 14/17: dpkg: Use a variable for each conffile pathname type

Author: Guillem Jover AuthorDate: Thu Apr 18 00:50:31 2024 +0200 dpkg: Use a variable for each conffile pathname type Instead of reusing two variables for various purposes, which makes the code hard to follow and is rather confusing, use a dedicated variable for each pathname type

[dpkg] 12/17: libdpkg: Add new file_getcwd() function

Author: Guillem Jover AuthorDate: Sat Apr 20 01:08:06 2024 +0200 libdpkg: Add new file_getcwd() function This new function abstracts the logic to retry getting the current working directory by growing a buffer so that the result can fit. --- lib/dpkg/file.c | 14

[dpkg] 11/17: libdpkg: Add new varbuf prefix and suffix handling functions

Author: Guillem Jover AuthorDate: Sat Apr 13 22:39:44 2024 +0200 libdpkg: Add new varbuf prefix and suffix handling functions These add checks for whether a prefix or a suffix is present in a varbuf, and functions to trim varbuf or repeated character prefixes. --- lib/dpkg/libdpkg.map

[dpkg] 01/17: test: Pass --check-level=exhaustive to cppcheck

Author: Guillem Jover AuthorDate: Fri Apr 26 09:48:42 2024 +0200 test: Pass --check-level=exhaustive to cppcheck Starting with cppcheck 2.14.0, it emits an information level tag that it is not analyzing all branches, which makes the test suite fail. Warned-by: cppcheck

[dpkg] 10/17: libdpkg: Add support for DPKG_NLS environment variable

Author: Guillem Jover AuthorDate: Wed Feb 28 04:34:21 2024 +0100 libdpkg: Add support for DPKG_NLS environment variable This implements the DPKG_NLS environment variable support that all other perl scripts already support, unifying the interface. This also covers the case where

[dpkg] 02/17: libdpkg: Change varbuf_get_str() to return "" instead of initializing it

Author: Guillem Jover AuthorDate: Fri Jul 21 00:11:30 2023 +0200 libdpkg: Change varbuf_get_str() to return "" instead of initializing it When we have a varbuf that has not been populated, growing its memory to simply add a NUL character is a waste. Instead simply return

[dpkg] 08/17: man: Add new libdpkg(7) manual page

Author: Guillem Jover AuthorDate: Thu Apr 11 00:52:34 2024 +0200 man: Add new libdpkg(7) manual page This is intended to document the library as a whole, give some overview of its usage, API stability and deprecations. The API documentation for this library

[dpkg] 04/17: Use varbuf_str() instead of direct access

Author: Guillem Jover AuthorDate: Fri Sep 1 03:53:31 2023 +0200 Use varbuf_str() instead of direct access Replace any access that can result in an uninitialized buf member, in case it only gets conditionally added strings to it. We might do a second pass for the remaining read-only

[dpkg] branch main updated (54a1a23d3 -> 51b9b23c4)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from 54a1a23d3 scripts/mk: Generate the _FOR_BUILD variant of each variable automatically new bfecc8f13 test: Pass --check-level=exhaustive to cppcheck new

[dpkg] 03/17: libdpkg: Rename varbuf_get_str() to varbuf_str()

Author: Guillem Jover AuthorDate: Fri Jul 21 00:11:30 2023 +0200 libdpkg: Rename varbuf_get_str() to varbuf_str() This shortens a bit the function name so that we can still fit it in long lines without having to wrap them around, and without losing any semantic meaning from its

[dpkg] branch 1.21.x updated (acb1c44ba -> 3f4f61a35)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch 1.21.x in repository dpkg. from acb1c44ba build: Build gitlab CI images for bookworm instead of sid add 3f4f61a35 po: Update Swedish translations No new revisions were added by this

Bug#1070027: fdisk: dependency problems prevent configuration of fdisk

Control: severity -1 normal Hi! On Mon, 2024-04-29 at 00:01:02 +0200, Chris Hofstaedtler wrote: > Control: reassign -1 dpkg > > * Vincent Lefevre [240428 22:33]: > > Package: fdisk > > Version: 2.40-8 > > Severity: serious > ... > > Setting up util-linux (2.40-8) ... > > fstrim.service is a

Re: Pondering sunsetting the Jenkins CI and repo for dpkg

Hi! On Thu, 2024-03-14 at 00:19:09 +0100, Alexandre Detiste wrote: > Le sam. 9 mars 2024 à 18:29, Guillem Jover a écrit : > > The grml.org project has been hosting Jenkins jobs for dpkg for a long > > time now, as the very initial CI we had. The jobs also generate a repo > &g

Bug#1069846: dpkg: dpkg-deb fails to build package w/ incomplete changelog entry

nother one refactoring the functions (which I had pending on doing). Thanks, Guillem From 93e7b2268fabc35d754ffe6389b5172b5917eb8c Mon Sep 17 00:00:00 2001 From: Guillem Jover Date: Thu, 25 Apr 2024 22:44:19 +0200 Subject: [PATCH] src: Check whether SOURCE_DATE_EPOCH is set before parsing it The dpkg-deb and dpkg-

Bug#1020533: Bug#825385: Bug#1020533: dpkg should use /var/lib/dpkg/arch to determine native arch when running chrootless

Hi! On Fri, 2023-06-16 at 16:29:17 +0200, Johannes Schauer Marin Rodrigues wrote: > Quoting Guillem Jover (2022-10-10 12:23:58) > > On Thu, 2022-09-22 at 22:13:34 +0200, Johannes Schauer Marin Rodrigues > > wrote: > > As mentioned on IRC, the problem here (and on #825385

Bug#1068483: Bug#882511: dpkg-buildpackage: should allow caller to force inclusion of source in buildinfo

Hi! On Wed, 2024-04-10 at 15:22:45 -0700, Vagrant Cascadian wrote: > On 2024-04-09, Guillem Jover wrote: > > I've now finished the change I had in that branch, which implements > > support so that dpkg-buildpackage can be passed a .dsc or a source-dir, > > and in the form

[dpkg] 05/05: scripts/mk: Generate the _FOR_BUILD variant of each variable automatically

-> main) Author: Nicolas Boulenguez AuthorDate: Sat Mar 9 17:21:53 2024 +0100 scripts/mk: Generate the _FOR_BUILD variant of each variable automatically Use a loop instead of duplicating the same variable with and without _FOR_BUILD suffix. Signed-off-by: Guillem Jo

[dpkg] 01/05: scripts/mk: Protect files against double inclusion

. [guil...@debian.org: - Rename variables. - Remove ifndef from endif comment. ] Signed-off-by: Guillem Jover --- scripts/mk/architecture.mk | 5 + scripts/mk/buildapi.mk | 5 + scripts/mk/buildflags.mk | 6 ++ scripts/mk/buildopts.mk| 5 + scripts/mk

[dpkg] 04/05: scripts/mk: Search once for parrallel= in DEB_BUILD_OPTIONS

Author: Nicolas Boulenguez AuthorDate: Tue Apr 9 05:31:16 2024 +0200 scripts/mk: Search once for parrallel= in DEB_BUILD_OPTIONS Previously each expansion of DEB_BUILD_OPTION_PARALLEL was parsing DEB_BUILD_OPTIONS. Signed-off-by: Guillem Jover --- scripts/mk/buildopts.mk | 8

[dpkg] 02/05: scripts/mk: Use filter instead of findstring

Author: Nicolas Boulenguez AuthorDate: Tue Apr 9 04:47:05 2024 +0200 scripts/mk: Use filter instead of findstring $(filter nostrip,$(DEB_BUILD_OPTIONS)) is slightly better than $(findstring) because the latter matches 'arduinostrip'. Signed-off-by: Guillem Jover --- scripts

[dpkg] branch main updated (f268c3733 -> 54a1a23d3)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from f268c3733 po: Update German man pages translation new 8966616fa scripts/mk: Protect files against double inclusion new c0dde13fd scripts/mk: Use filter

[dpkg] 03/05: scripts/mk: Use explicit test of $(origin) instead of ?=

Author: Nicolas Boulenguez AuthorDate: Tue Apr 9 04:56:26 2024 +0200 scripts/mk: Use explicit test of $(origin) instead of ?= This makes the intent of each stanza more readable (first ensure that TOOL is set, then TOOL_FOR_BUILD). Signed-off-by: Guillem Jover --- scripts/mk

Bug#872381: dpkg-dev: optimize Makefile snippets for debian/rules

e speed > gain > architecture.mk 30 times faster (probably no gain under dpkg-buildpackage) > buildflags.mk 20 times faster > pkg-info.mk 4 times faster > buildtools.mk20% faster > > Guillem Jover > > I've left this one out f

Bug#1068483: Bug#882511: dpkg-buildpackage: should allow caller to force inclusion of source in buildinfo

nstead > of > the attached patch. The problem with this solution is that it is prone do accidental use, as it is very easy for a user to unknowingly have recreated the sources from a locally extracted tree (be that modified or not). On Sat, 2024-04-06 at 02:57:40 +0200, Guille

Bug#1068483: dpkg-genbuildinfo: Should buildinfo files copy the hash of the source package?

Hi! On Sat, 2024-04-06 at 02:56:02 +0300, Adrian Bunk wrote: > Package: dpkg-dev > Version: 1.22.6 > Severity: normal > X-Debbugs-Cc: reproducible-bui...@lists.alioth.debian.org > A thought I already wrote in a recent debian-devel discussion: > > In theory source package filenames should be

Bug#1067087: dpkg-buildflags: Enable frame pointers by default

Control: tags -1 moreinfo Control: retitle -1 dpkg-buildflags: Enable frame pointers by default Hi! On Mon, 2024-03-18 at 09:56:38 +0100, Kurt Roeckx wrote: > Source: gcc-14 > Severity: wishlist > Please consider enabling frame pointers on 64 bit arches. > See: >

[dpkg] branch main updated: man: Document DEB_BUILD_ARCH and DEB_HOST_ARCH usage in commands

by this push: new 3e430c1f2 man: Document DEB_BUILD_ARCH and DEB_HOST_ARCH usage in commands 3e430c1f2 is described below commit 3e430c1f2fdf4b27ca4050ea4cb53dfec7206d76 (HEAD -> main) Author: Guillem Jover AuthorDate: Tue Apr 2 02:40:48 2024 +0200 man: Document DEB_BUILD_A

[dpkg] 02/02: build: Add support to track release VCS commit id

-> main) Author: Guillem Jover AuthorDate: Mon Apr 1 16:43:48 2024 +0200 build: Add support to track release VCS commit id This adds a provenance trail from the upstream VCS commit the tarball was supposedly generated from, so that it can be traced by interested downstre

[dpkg] 01/02: build: Print the release version at the end of configure

Author: Guillem Jover AuthorDate: Mon Apr 1 16:44:57 2024 +0200 build: Print the release version at the end of configure This should give a bit more traceability during builds by downstreams. --- configure.ac | 3 +++ 1 file changed, 3 insertions(+) diff --git a/configure.ac b

[dpkg] branch main updated (4d9214444 -> 54a6ad9db)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from 4d921 po: Update German man pages translation new 00707c50e build: Print the release version at the end of configure new 54a6ad9db build: Add

Bug#1067876: dpkg allows installation of malformed .deb packages resulting in corrupt .list file

Hi! On Thu, 2024-03-28 at 09:54:36 +, Grueninger, Tobias wrote: > Package: dpkg > Version: 1.20.12 > Severity: wishlist > In our case we did install a package from a 3rd party which apparently > uses a non-conforming .deb package building tool (unknown to us) > resulting in a malformed

[dpkg] branch main updated: build: Partially revert the sanitizer for some functional tests in CI

by this push: new 2c55113b6 build: Partially revert the sanitizer for some functional tests in CI 2c55113b6 is described below commit 2c55113b69485a1f776f878a7bba777511354397 (HEAD -> main) Author: Guillem Jover AuthorDate: Wed Mar 27 03:37:16 2024 +0100 build: Partially rev

[dpkg] 05/06: libdpkg: Handle readlink() failures in file_readlink()

Author: Guillem Jover AuthorDate: Tue Mar 26 04:01:30 2024 +0100 libdpkg: Handle readlink() failures in file_readlink() We should not use the linksize to update the varbuf if readlink() failed, as that implies a negative number. Reported-by: coverity --- lib/dpkg/file.c | 3

[dpkg] 06/06: Dpkg::Vendor::Ubuntu: Use -fcf-protection=none instead of -fno-cf-protection

-> main) Author: Matthias Klose AuthorDate: Wed Mar 27 02:43:39 2024 +0100 Dpkg::Vendor::Ubuntu: Use -fcf-protection=none instead of -fno-cf-protection The latter is not known to gcc. Signed-off-by: Guillem Jover --- scripts/Dpkg/Vendor/Ubuntu.pm | 2 +- 1 file changed

[dpkg] 04/06: Dpkg::BuildDriver: Refactor build driver out of dpkg-buildpackage

Author: Guillem Jover AuthorDate: Sun Mar 24 05:44:54 2024 +0100 Dpkg::BuildDriver: Refactor build driver out of dpkg-buildpackage Add a new module with the logic to drive debian/rules. This also adds a new experimental field that can be used to select the build driver to use

[dpkg] 03/06: dpkg-buildpackage: Refactor build target hook execution

Author: Guillem Jover AuthorDate: Sat Feb 10 14:20:28 2024 +0100 dpkg-buildpackage: Refactor build target hook execution This unifies the handling with the other hook execution. It has a slight behavior change, where the DPKG_BUILDPACKAGE_HOOK_BUILD_TARGET is now set even when

[dpkg] 02/06: dpkg-buildpackage: Fix the debian/rules executable check to respect -R

Author: Guillem Jover AuthorDate: Sat Feb 3 17:58:04 2024 +0100 dpkg-buildpackage: Fix the debian/rules executable check to respect -R When the user has specified a different rules file via the -R option, we should try to use that file for the check that makes sure it is executable

[dpkg] 01/06: dpkg-buildpackage: Remove fallback handling for missing required targets

Author: Guillem Jover AuthorDate: Sat Feb 10 13:43:24 2024 +0100 dpkg-buildpackage: Remove fallback handling for missing required targets The fallback code that was auto-detecting for potentially missing build-arch and build-indep targets has been there since 2012, and has

[dpkg] branch main updated (7b6e42a13 -> 78a5e8e24)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from 7b6e42a13 libcompat: Include missing in strnlen module new f4a75f138 dpkg-buildpackage: Remove fallback handling for missing required targets new

Bug#1059150: No longer works with signing subkeys

Hi! On Wed, 2024-03-20 at 19:05:59 +, Steve McIntyre wrote: > On Wed, Dec 20, 2023 at 11:59:31PM +0100, Guillem Jover wrote: > >On Wed, 2023-12-20 at 15:30:24 +, Steve McIntyre wrote: > >> Package: debsig-verify > >> Version: 0.23+b2 > >> Severity:

Bug#1059150: No longer works with signing subkeys

Hi! On Wed, 2024-03-20 at 18:00:30 +, Steve McIntyre wrote: > On Wed, Mar 20, 2024 at 05:18:08PM +, Steve McIntyre wrote: > >Sorry, I've been swamped with other stuff then ill for the last week > >or so. Looking now... No worries, hope you are doing well now! :) > And I can confirm that

Bug#1065429: dpkg -s: spurious error "dpkg-query: error: --status needs a valid package name"

Hi! On Thu, 2024-03-07 at 04:25:03 +0100, Vincent Lefevre wrote: > On 2024-03-07 03:34:08 +0100, Guillem Jover wrote: > > > "apt-cache show libc6-dev" also lists this package. > > > > apt behaves differently, this has been also a known discrepancy, bu

[dpkg] 05/09: build: Add missing space before backslash

Author: Guillem Jover AuthorDate: Tue Nov 21 22:38:36 2023 +0100 build: Add missing space before backslash --- lib/dpkg/Makefile.am | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/dpkg/Makefile.am b/lib/dpkg/Makefile.am index b0243b157..8c72cbad1 100644 --- a/lib/dpkg

[dpkg] 06/09: build: Unconditionally include

Author: Guillem Jover AuthorDate: Wed Mar 20 00:30:59 2024 +0100 build: Unconditionally include We did this for libdpkg in commit 3516d105a427ad95c4d6bacb96030e4267c842fd, which was unifying the assumption with the rest of the code base, so there is no point in conditionally

[dpkg] 07/09: build: Do not check for memcpy()

Author: Guillem Jover AuthorDate: Wed Mar 20 00:40:49 2024 +0100 build: Do not check for memcpy() This function should be present already in C89 and we require C99 now. In addition it was causing configure failures when building with «-Werror». Reported-by: Simon Richter

[dpkg] 09/09: libcompat: Include missing in strnlen module

-> main) Author: Guillem Jover AuthorDate: Wed Mar 20 00:49:17 2024 +0100 libcompat: Include missing in strnlen module We need the header for the memchr() prototype. Reported-by: Simon Richter --- lib/compat/strnlen.c | 2 ++ 1 file changed, 2 insertions(+) diff --

[dpkg] 08/09: build: Do not check for functions used unconditionally

Author: Guillem Jover AuthorDate: Wed Mar 20 00:42:49 2024 +0100 build: Do not check for functions used unconditionally We expect these functions to be present as we are using them unconditionally, so there is no point in checking for them. --- configure.ac | 2 -- 1 file changed, 2

[dpkg] 03/09: doc: Add missing full stop to end of sentence

Author: Guillem Jover AuthorDate: Sun Mar 17 22:54:04 2024 +0100 doc: Add missing full stop to end of sentence --- tests/README | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/README b/tests/README index 3da99ab00..cea05d0be 100644 --- a/tests/README +++ b/tests/README

[dpkg] 02/09: test: Do not fail the functional test suite due to memory leaks

Author: Guillem Jover AuthorDate: Sun Mar 17 18:40:29 2024 +0100 test: Do not fail the functional test suite due to memory leaks The codebase still has many static variables, and variables used during error conditions that are not cleaned up on exit or error, either because

[dpkg] 04/09: build: Re-enable the sanitizer for functional tests in CI

Author: Guillem Jover AuthorDate: Sun Mar 17 19:52:25 2024 +0100 build: Re-enable the sanitizer for functional tests in CI Now that we are ignoring leak sanitizer errors, we can enable the sanitizer checks for all the rest of issues that can be detected. At the same time we

[dpkg] branch main updated (cedd58200 -> 7b6e42a13)

This is an automated email from the git hooks/post-receive script. guillem pushed a change to branch main in repository dpkg. from cedd58200 po: Update German man pages translation new 5cdc783e5 libdpkg: Factor out filesystem database file loading into new function new 0d3527afd

Re: [PATCH] Valgrind support for test suite

support for test suite [guil...@debian.org: - Only export DPKG_WRAPPER in the memtest target. - Improve code comment. - Make the memtest target depend on test. ] Signed-off-by: Guillem Jover --- tests/Makefile | 3 +++ tests/Test.mk | 15 +-- 2 files changed, 12 insertions(+), 6

Re: [PATCH] Factor out common code for reloading database files

Hi! On Mon, 2024-03-18 at 13:14:14 +0900, Simon Richter wrote: > On 3/18/24 01:28, Guillem Jover wrote: > > Attached is what I had, which I've polished a bit more now. Will > > include in my next push. > > Nice, will rebase my branch on top of that. The new file probabl

Re: [PATCH] cleanup: Pull in for memchr

Hi! On Tue, 2024-03-19 at 23:48:00 +0900, Simon Richter wrote: > diff --git a/lib/compat/strnlen.c b/lib/compat/strnlen.c > index d02bb4bbd..6a7eb0454 100644 > --- a/lib/compat/strnlen.c > +++ b/lib/compat/strnlen.c > @@ -22,6 +22,10 @@ > > #include "compat.h" > > +#ifdef HAVE_STRING_H > +#

Re: [PATCH] cleanup: make memcpy check -Werror safe

Hi! Thanks! I ended up just removing the check for this function. Now in git main. Regards, Guillem

Re: [PATCH] Factor out common code for reloading database files

defc49ccbf2e626155e77ba77a141fa9033473a0 Mon Sep 17 00:00:00 2001 From: Guillem Jover Date: Sun, 17 Mar 2024 15:02:45 +0100 Subject: [PATCH] libdpkg: Factor out filesystem database file loading into new function This code is duplicated on several places that load filesystem databases, refactor it into a new function

Re: Another usrmerge complication

Hi! On Sun, 2024-03-17 at 11:23:28 +0900, Simon Richter wrote: > because life isn't hard enough as it is: When /bin is a symlink to usr/bin, > and I install two packages, where one installs /bin/foo and the other > installs /usr/bin/foo, then, if both are installed in the same dpkg > invocation,

Bug#1066952: Defaulting to -Werror=implicit-function-declaration might not be a good idea

Hi! On Sat, 2024-03-16 at 02:18:22 +0200, Adrian Bunk wrote: > Package: dpkg-dev > Version: 1.22.6 > Severity: serious > X-Debbugs-Cc: Steve Langasek > 3. Code that does still compile with -Werror=implicit-function-declaration >but behaves differently, e.g. due to failing autoconf tests

[dpkg] 01/03: po: Regenerate .pot files and merge .po files with them

Author: Guillem Jover AuthorDate: Sun Mar 10 20:21:23 2024 +0100 po: Regenerate .pot files and merge .po files with them --- dselect/po/bs.po| 2 +- dselect/po/ca.po| 2 +- dselect/po/cs.po| 2 +- dselect/po/da.po| 2 +- dselect/po/de.po| 2 +- dselect

  1   2   3   4   5   6   7   8   9   10   >