Bug#1074577: gnat ftbfs with glibc from experimental
Source: gcc-13 Followup-For: Bug #1074577 Matthias Klose: > Simon Chopin came up with a conditional to work with both variants [before and after glibc introduces __USE_TIME64_REDIRECTS] > #if defined(__USE_TIME64_REDIRECTS) || (__TIMESIZE == 32 && __USE_TIME_BITS64) I have forwarded this upstream as version 11. This changes the libgnat-13 sources (System.OS_Constants). Most Ada packages will require a bin-NMU after you upload gcc-13, ideally before people start filling FTBS bug reports. I take the opportunity to apply some postponed changes. I do not commit them because I have tested no build, but the only changes since the (tested) version 10 are * the line above, tested by Simon Chopin in Ubuntu * unapplying small style changes reviewed by upstream From: Nicolas Boulenguez Subject: [PATCH] Ada: update patches for PR114065 (time_64) to v11 The fix 2bacf86d for #1074577 by Simon Chopin at https://bugs.launchpad.net/ubuntu/+source/gcc-13/+bug/2071605 requires a rebuild of all Ada libraries, so we might as well * update all patches to version 11 * apply the parts fixing bugs, but not the style suggestions --- ...ersions-with-C-struct-timeval-from-GN.diff | 143 - ...ersions-with-C-struct-timespec-from-A.diff | 167 - ...ersions-with-C-time_t-from-System.OS_.diff | 79 --- ...imeval-and-timespec-definitions-and-c.diff | 604 ++ ...-unneeded-x32-variant-of-System.Linux.diff | 146 - ...ed-posix2008-variant-of-System.Parame.diff | 229 --- ...ed-darwin-solaris-x32-variants-of-Sys.diff | 472 -- ...sleep-from-System.OS_Primitives.Timed.diff | 74 --- ...its-time-functions-from-GNU-libc-when.diff | 74 ++- debian/rules.patch| 7 - 10 files changed, 520 insertions(+), 1475 deletions(-) delete mode 100644 debian/patches/0001-Ada-remove-conversions-with-C-struct-timeval-from-GN.diff delete mode 100644 debian/patches/0002-Ada-remove-conversions-with-C-struct-timespec-from-A.diff delete mode 100644 debian/patches/0003-Ada-remove-conversions-with-C-time_t-from-System.OS_.diff delete mode 100644 debian/patches/0005-Ada-drop-unneeded-x32-variant-of-System.Linux.diff delete mode 100644 debian/patches/0006-Ada-drop-unneeded-posix2008-variant-of-System.Parame.diff delete mode 100644 debian/patches/0007-Ada-drop-unneeded-darwin-solaris-x32-variants-of-Sys.diff delete mode 100644 debian/patches/0008-Ada-import-nanosleep-from-System.OS_Primitives.Timed.diff diff --git a/debian/patches/0001-Ada-remove-conversions-with-C-struct-timeval-from-GN.diff b/debian/patches/0001-Ada-remove-conversions-with-C-struct-timeval-from-GN.diff deleted file mode 100644 index d57d7ec7.. --- a/debian/patches/0001-Ada-remove-conversions-with-C-struct-timeval-from-GN.diff +++ /dev/null @@ -1,143 +0,0 @@ -From bedb7553c420da59938eacb115fd9384e54ceae0 Mon Sep 17 00:00:00 2001 -From: Nicolas Boulenguez -Date: Fri, 5 Apr 2024 16:51:54 +0200 -Subject: [PATCH 1/9] Ada: remove conversions with C struct timeval from - GNAT.Calendar - - gcc/ada/doc/gnat_rm/the_gnat_library.rst | 2 - - gcc/ada/libgnat/g-calend.adb | 58 - gcc/ada/libgnat/g-calend.ads | 18 - 3 files changed, 78 deletions(-) - -diff --git a/src/gcc/ada/doc/gnat_rm/the_gnat_library.rst b/src/gcc/ada/doc/gnat_rm/the_gnat_library.rst -index 3aae70a..bcec49f 100644 a/src/gcc/ada/doc/gnat_rm/the_gnat_library.rst -+++ b/src/gcc/ada/doc/gnat_rm/the_gnat_library.rst -@@ -674,8 +674,6 @@ Machine-specific implementations are available in some cases. - - Extends the facilities provided by ``Ada.Calendar`` to include handling - of days of the week, an extended ``Split`` and ``Time_Of`` capability. --Also provides conversion of ``Ada.Calendar.Time`` values to and from the --C ``timeval`` format. - - .. _`GNAT.Calendar.Time_IO_(g-catiio.ads)`: - -diff --git a/src/gcc/ada/libgnat/g-calend.adb b/src/gcc/ada/libgnat/g-calend.adb -index 0a98eb2..e0d34f5 100644 a/src/gcc/ada/libgnat/g-calend.adb -+++ b/src/gcc/ada/libgnat/g-calend.adb -@@ -29,11 +29,8 @@ - -- -- - -- - --with Interfaces.C.Extensions; -- - package body GNAT.Calendar is -use Ada.Calendar; -- use Interfaces; - -- --- Day_In_Year -- -@@ -328,61 +325,6 @@ package body GNAT.Calendar is -Time_Zone=> 0); -end Time_Of_At_Locale; - -- - -- -- To_Duration -- -- - -- -- function To_Duration (T : not null access timeval) return Duration is -- -- procedure timeval_to_duration --(T: not null access timeval; -- sec : not null access C.Extensions.long_long; -- usec : not null access C.long); -- pragma Import (C, timeval_to_duration, "__gnat_timeval_to_duration"); -- --
Bug#1067285: whitakers-words in Debian, gnat-13 transition
Hello. The package needs to build with gnat-13 in order to be part of next Debian release. Are you planning a new upload in the near future? Else, are you OK with a non maintainer upload fixing this specific bug?
Bug#1067320: topal: FTBFS: debian/rules: debian_packaging.mk: No such file or directory
Source: topal Followup-For: Bug #1067320 Control: tag -1 + patch Hello. Attachment 002 below fixes this bug. Would you be OK with a non maintainer upload? The other attachments are unrelated sugestions. Would you be OK with a salsa.debian.org/debian/topal git repository? PATH 1/10 updates the upstream part to version 82. >From 31c2f14e91e2a01c75eb1309f17ea540ffb80571 Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Sat, 3 Dec 2022 18:56:23 +0100 Subject: [PATCH 02/10] Switch to dh-ada-library >= 8.2 for packaging.mk --- debian/control | 1 + debian/rules | 3 +-- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/debian/control b/debian/control index be7984d..53206e5 100644 --- a/debian/control +++ b/debian/control @@ -4,6 +4,7 @@ Priority: optional Maintainer: Phil Brooke Build-Depends: debhelper-compat (= 13), + dh-ada-library (>= 8.2), gnat (>= 11), libreadline-dev, texlive, diff --git a/debian/rules b/debian/rules index cdd07e4..d3e38d5 100755 --- a/debian/rules +++ b/debian/rules @@ -4,8 +4,7 @@ DEB_BUILD_MAINT_OPTIONS := hardening=+all include /usr/share/dpkg/buildflags.mk -include /usr/share/dpkg/buildopts.mk -include /usr/share/ada/debian_packaging.mk +include /usr/share/ada/packaging.mk # Compile Ada and C with the same compiler. CC := gnatgcc -- 2.39.2 >From a9e088d51c8f95a9e7cd60bee7e31f9e167834ed Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Sat, 3 Dec 2022 18:57:35 +0100 Subject: [PATCH 03/10] Set CC from gnat version without the deprecated gnatgcc symbolic link gcc-$MAJOR is also specific to Debian, but not to Ada. --- debian/rules | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/debian/rules b/debian/rules index d3e38d5..b29f0ea 100755 --- a/debian/rules +++ b/debian/rules @@ -7,7 +7,7 @@ include /usr/share/dpkg/buildflags.mk include /usr/share/ada/packaging.mk # Compile Ada and C with the same compiler. -CC := gnatgcc +CC := gcc-$(DEB_GNAT_VERSION) # Upstream Makefile insists on rebuilding everything everytime. # SOURCE_DATE_EPOCH (set by debhelper) and the -m gnatmake option may -- 2.39.2 >From 6a3416662d4142d461f6c59bdaca4c3e495f59ff Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Wed, 20 Jul 2022 14:56:41 +0200 Subject: [PATCH 04/10] Drop full texlive from build dependencies It was now redundant with more specific dependencies. --- debian/control | 1 - 1 file changed, 1 deletion(-) diff --git a/debian/control b/debian/control index 53206e5..5669b2f 100644 --- a/debian/control +++ b/debian/control @@ -7,7 +7,6 @@ Build-Depends: dh-ada-library (>= 8.2), gnat (>= 11), libreadline-dev, - texlive, texlive-latex-base, texlive-latex-extra, texlive-fonts-recommended, -- 2.39.2 >From 2946bec7938ae412ac7cf632a2f2982a6b79355d Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Wed, 20 Jul 2022 14:57:05 +0200 Subject: [PATCH 05/10] Sort build dependencies Dpkg supports the extra comma exactly for this purpose. --- debian/control | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/debian/control b/debian/control index 5669b2f..00262aa 100644 --- a/debian/control +++ b/debian/control @@ -7,10 +7,10 @@ Build-Depends: dh-ada-library (>= 8.2), gnat (>= 11), libreadline-dev, + texlive-fonts-extra, + texlive-fonts-recommended, texlive-latex-base, texlive-latex-extra, - texlive-fonts-recommended, - texlive-fonts-extra Standards-Version: 4.6.1 Rules-Requires-Root: no Homepage: https://www.zircon.org.uk/topal/ -- 2.39.2 >From acb27a97b77357e5a17617c778119151161468de Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Sat, 3 Dec 2022 19:18:57 +0100 Subject: [PATCH 06/10] Remove some trailing whitespaces --- debian/changelog | 1 - debian/rules | 1 - 2 files changed, 2 deletions(-) diff --git a/debian/changelog b/debian/changelog index 03c854b..ba63f7a 100644 --- a/debian/changelog +++ b/debian/changelog @@ -290,4 +290,3 @@ topal (0.6.4-1) unstable; urgency=low * Initial Release (closes: #143319). -- Phil Brooke Sun, 21 Apr 2002 14:29:05 + - diff --git a/debian/rules b/debian/rules index b29f0ea..c8539a8 100755 --- a/debian/rules +++ b/debian/rules @@ -37,4 +37,3 @@ override_dh_auto_install: .PHONY: override_dh_installchangelogs override_dh_installchangelogs: dh_installchangelogs Changelog.html - -- 2.39.2 >From 6934d431d7d4efde5a7a2e2095e879d6488a083d Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Sun, 27 Aug 2023 14:03:48 +0200 Subject: [PATCH 07/10] Bump Standards-Version --- debian/control | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/debian/control b/debian/control index 00262aa..abc38ee 100644 --- a/debian/control +++ b/debian/control @@ -11,7 +11,7 @@ Build-Depends: texlive-fonts-recommended, texlive-latex-base, texlive-latex-extra, -Standards-Version: 4.6.1 +Standards-Version: 4.7.0 Rules-Requires-Root: no Ho
Bug#1069984: alire: Build-depends on NBS package libgnatcoll21-dev
Source: alire Followup-For: Bug #1069984 Hello. This bug is already fixed in 1.2.1-1.1. The ideal way to close a bug is usually in debian/changelog, but this bug was open after its fix has been uploaded to experimental, so you should probably close it as described at https://www.debian.org/Bugs/Developer#closing with a mail to 1069984-d...@bugs.debian.org starting with Version: 1.2.1-1.1 The 'debian/ 1.2.1-2' tag is only informative, it triggers no actual upload. However, its current contents are incorrect. debian/changelog contains the same version twice. Please fix this, then I will sponsor an upload and you will tag the uploaded state as 'debian/1.2.1-2+really'.
Bug#1067285: whitakers-words: FTBFS: make[2]: *** [Makefile:40: commands] Error 4
Source: whitakers-words Followup-For: Bug #1067285 Hello. This failure is caused by new compiler warnings, either because of the switch to gnat-13 or because of changes in Debian options. Anyway, I suggest to add the following two lines in debian/rules, somewhere before the inclusion of /usr/share/*.mk. # Disable the -gnatwe upstream flags during Debian builds. DEB_ADAFLAGS_MAINT_APPEND := -gnatwn
Bug#1027469: FTBFS fixed by 0.2020.10.27-1.3
Version: 0.2020.10.27-1.3 Hello. I have uploaded an NMU without delay, because * this blocks the gnat-12 transition * he has already reviewed these changes in the VCS while we were preparing the next upload.
Bug#1026828: downgrade the severity
Control: severity 1026828 normal Control: tags 1026828 - ftbfs Control: retitle 1026828 random print error for floats on i386 Version 1.5-3 transforms the error into a warning (with more debugging info). The original bug does not deserve a 'serious' severity.
Bug#1026828: libgmpada: FTBFS on i386: raised ADA.ASSERTIONS.ASSERTION_ERROR : demo.adb:119
Source: libgmpada Version: 1.5-1 Severity: serious Tags: ftbfs Justification: fails to build on a supported architecture A post-build test fails on i386 buildds: https://buildd.debian.org/status/fetch.php?pkg=libgmpada=i386=1.5-1=1671621582=0 https://buildd.debian.org/status/fetch.php?pkg=libgmpada=i386=1.5-1=1661971646=0 but succeeds on an i386 porterbox.
Bug#1025733: merge 1020018 1025733
forcemerge 1020018 1025733 thanks This is a duplicate. 1020018 is closed by libxmlada/23.0.0-1 in experimental.
Bug#936869: RM: libgnatcoll-python -- ROM; NPOASR; temporary package not needed anymore
Package: ftp.debian.org
Severity: normal
Control: unblock 936624 by 936869
Hello.
Please remove these packages from unstable:
libgnatcoll-python:src
libgnatcoll-python19
libgnatcoll-python19-dev
I am the maintainer of libgnatcoll-bindings and libgnatcoll-python.
The normal situation is that
the libgnatcoll-bindings source package produces
libgnatcoll-${COMPONENT}{SOVERSION} and
libgnatcoll-${COMPONENT}{ALIVERSION}-dev
for various COMPONENTs including python.
A few years ago, the python component was ready for both python2&3,
but its only reverse dependency gnat-gps was requiring python2.
As a temporary measure, the separate libgnatcoll-python source package
was introduced. This allowed the other components to be updated in
testing, while the python(2) component and gnat-gps were RC-buggy in
unstable (#936869 and #936624 respectively).
Nowadays
* gnat-gps has been RC-buggy for a while.
It depends on libgnatcoll-python18 that no one is building anymore.
* If/when we manage to rebuild gnat-gps, latest upstream version is
ready for python3. Hence the unblock command above.
There is no need for a separate package anymore, so
libgnatcoll-bindings builds the python(3) component again.
libgnatcoll-python3.1
libgnatcoll-python3.1-dev
These fully replace the removed packages.
Thanks.
Bug#1008152: gprbuild: FTBFS during binary-indep build
Source: gprbuild Followup-For: Bug #1008152 Control: severity -1 normal
Bug#997728: libgnatcoll-db: FTBFS: raised CONSTRAINT_ERROR : uintp.adb:2002 overflow check failed
Source: libgnatcoll-db Followup-For: Bug #997728 Control: severity -1 normal Control: reassign -1 gnat-10 Control: tags -1 - bookworm ftbfs sid Control: merge -1 954681 Hello. This is most probably caused by the SOURCE_DATE_EPOCH patch applied by Debian when building GNAT from GCC sources. It has been fixed in most cases (especially, on buildds), but apparently not for all SOURCE_DATE_EPOCH values (see #954681),
Bug#1008982: adacontrol: depends on unavailable asis
Source: adacontrol Severity: serious Tags: upstream ftbfs Justification: Policy 2.2.1 ASIS will soon be removed from Debian. Adacontrol depends on ASIS and should also be removed.
Bug#1008981: libaws: depends on unavailable asis
Source: libaws Severity: serious Tags: upstream ftbfs Justification: Policy 2.2.1 ASIS will soon be removed from Debian. AWS depends on ASIS and should also be removed. Next AWS version will depend on langkit instead, but langkit is not packaged yet.
Bug#1008980: adabrowse: depends on unavailable asis
Package: adabrowse Severity: serious Tags: upstream ftbfs Justification: Policy 2.2.1 ASIS will soon be removed from Debian. adabrowse depends on ASIS and should also be removed.
Bug#1008979: asis: depends on gnat-util, almost orphaned upstream
Source: asis Severity: serious Tags: upstream ftbfs Justification: Policy 2.2.1 gcc-11 has stopped building the libgnat-util library (again). This issue alone could be circumvented by Debian efforts (and has been for more than a decade), but the only consumer left is ASIS. AdaCore does not release ASIS publicly anymore. https://www.adalog.fr/en/adacontrol.html#download gives much more details and describes possible work-arounds. For example, ASIS in Debian currently builds with GCC-10 from its last AdaCore public release (2019), but this requires libgnat-util. For now, ASIS must be removed from testing so that other Ada packages can migrate to GCC-11.
Bug#1008152: gprbuild: FTBFS during binary-indep build
Source: gprbuild Followup-For: Bug #1008152 Control: tags -1 + unreproducible Hello. For unrelated reasons, I have updated gprbuild in experimental. The -indep build seems to succeed on a build machine. https://buildd.debian.org/status/fetch.php?pkg=gprbuild=all=2022.0.0-2=1648992585=0 Until we find a way to reproduce the failure, I suggest to lower the severity to important or normal.
Bug#1008152: gprbuild: FTBFS during binary-indep build
Source: gprbuild Followup-For: Bug #1008152 Hello. In a local pbuilder chroot, gprbuild/2022.0.0-1 seems to build architecture-independant packages successfully. I have tried the same build with 2021.0.0.0778b109-5, with tex-gyre installed manually as a work-around for #997415. The build succeeds.
Bug#1005669: binutils-or1k-elf: FTBFS: autoreconf2.69: 'configure.ac' or 'configure.in' is required
Source: binutils-or1k-elf Followup-For: Bug #1005669 > > Apparently, a configure.ac has been forgotten alongside the generated > > configure. > > Commit dd96117a5e1da80ad1f977ea8abe39acc9f7e7cb works around the > > issue. > Hmm - I don't see the commit you mention above - did you perhaps forget > to push? Otherwise please provide me a URI so I know exactly what you > are talking about. I don’t know what is wrong. I can see it at: https://salsa.debian.org/debian/binutils-or1k-elf/-/commit/dd96117a5e1da80ad1f977ea8abe39acc9f7e7cb I am attaching the patch just in case. > Also, I can suggest that you post the information to > 1005...@bugs.debian.org instead of discretely here. You are right, of course. Done. Thanks. >From dd96117a5e1da80ad1f977ea8abe39acc9f7e7cb Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez Date: Sun, 13 Feb 2022 19:07:19 +0100 Subject: debian/autoreconf: skip etc/ in 2.38-1, configure.ac is missing diff --git a/debian/autoreconf b/debian/autoreconf index f2daa59..88e8205 100755 --- a/debian/autoreconf +++ b/debian/autoreconf @@ -12,6 +12,9 @@ set -C -e -u # Exclude zlib, which fails (binutils/2.35.2 autoconf/2.71). # It is ignored in favor of the Debian package anyway. +# Exclude etc, which contains etc/configure but not its source +# (since 2.38-1). + echo src -dirname src/*/configure | grep -v -e zlib -e libiberty +dirname src/*/configure | grep -v -e zlib -e libiberty -e etc
Bug#997414: debomatic: FTBFS: '! LaTeX Error: File `tgtermes.sty' not found.'
Package: debomatic Followup-For: Bug #997414 Control: tag -1 patch Hello. This issue is fixed by tex-gyre in Build-Depends.
Bug#981422: xdg-utils: autopkgtest failure: make: dh: No such file or directory
Source: xdg-utils
Followup-For: Bug #981422
Hello
The attached commit suggests a third option.
If you call the scripts with an explicit interpreter,
they need no execution rights in the first place.
>From 5a48d85ad66d60e9be0085e98fdc6ab830cf2e9f Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Wed, 31 Mar 2021 17:05:29 +0200
Subject: autotests: in test-lib.sh, call tty{on,off} with an explicit
interpreter
Explicit interpreters are usually discouraged so the calling code is
not broken by changes in the implementation, but in this case,
autotests/tty{on,off} are private helpers for autotests/test-lib.sh.
With an explicit interpreter, the scripts do not need execution rights.
Closes: #981422.
diff --git a/autotests/test-lib.sh b/autotests/test-lib.sh
index 87cab44..a16eb93 100644
--- a/autotests/test-lib.sh
+++ b/autotests/test-lib.sh
@@ -200,9 +200,9 @@ run() {
local trace=
if [ "$DETACH" = 0 ]; then
-trace="./ttyon"
+trace="/usr/bin/perl ttyon"
elif [ "$DETACH" = 1 ]; then
-trace="./ttyoff"
+trace="/usr/bin/perl ttyoff"
fi
trace="$trace ${SHELL:-/bin/sh}"
if [ "$TRACE" = 1 ]; then
diff --git a/autotests/ttyoff b/autotests/ttyoff
old mode 100755
new mode 100644
diff --git a/autotests/ttyon b/autotests/ttyon
old mode 100755
new mode 100644
diff --git a/debian/rules b/debian/rules
index 50c1d53..59b7c05 100755
--- a/debian/rules
+++ b/debian/rules
@@ -26,8 +26,6 @@ override_dh_auto_clean:
override_dh_auto_test:
ifeq (,$(filter nocheck,$(DEB_BUILD_OPTIONS)))
-# Fix perms of new files from the patch.
- chmod +x autotests/ttyon autotests/ttyoff
# Use automated tests that do not require a user action.
make autotest
endif
diff --git a/debian/tests/entry b/debian/tests/entry
index 4c1e834..7ba19b7 100755
--- a/debian/tests/entry
+++ b/debian/tests/entry
@@ -13,6 +13,5 @@ fi
BASH_XTRACEFD=1
set -x
-debian/rules patch
./configure
make autotest SHELL="${1:-/bin/sh}"
Bug#985512: liblingot-dev: broken symlink: /usr/lib/x86_64-linux-gnu/liblingot.so -> liblingot.so.0.0.0
Package: liblingot-dev Followup-For: Bug #985512 Control: tags -1 pending The VCS contains a trivial fix. https://salsa.debian.org/debian/lingot/-/commit/5ef26d260e2f6858e3226223290e052237b92e00
Bug#976354: git - tests fail with dash
Package: git Followup-For: Bug #976354 The test scripts include test-lib.sh, which deliberately assumes bash. # checkbashisms test-lib.sh possible bashism in test-lib.sh line 381 ($BASH_SOMETHING): if test -n "$BASH_VERSION" && eval ' This part happens to work for other shells, but the comments around it suggest that the intent is only to check the bash version. possible bashism in test-lib.sh line 1491 (builtin): builtin pwd -W The 'builtin' builtin only exists in bash. So I suggest to apply Bastian’s suggestion: # sed -i '1 s_^#!/bin/sh$_#!/bin/bash_' t/t*.sh
Bug#971018: libgnatcoll-db FTBFS on mips(64)el with assembler message: branch out of range
Hello. Here is the full history, just in case. mipsel mips64el 18-4 OK OK 19.2-1 fails OK 19.2-2 fails OK 19.2-3 -mxgot OK OK 20.0-1 -mxgot fails fails 21.0.0-1-mxgot fails-mxgot ? 21.0.0-2-mxgot fails-mxgot OK 21.0.0-3-mxgot fails-mxgot OK 21.0.0-4-mxgot fails-mxgot OK 21.0.0-5 fails-mxgot OK
Bug#975818: libgtkada: FTBFS: AttributeError: no attribute 'getchildren
Package: src:libgtkada Followup-For: Bug #975818 Control: tags -1 + pending The issue is caused by a changed in the python standard library. Version 21.0.0.785f3cf4-1 in experimental contains a fix, which will migrate during the gnat-10 transition.
Bug#972481: ruby-pygments.rb: FTBFS, tests failure (due to new pygments)
Package: src:ruby-pygments.rb
Followup-For: Bug #972481
Control: tags -1 + patch
Hello.
The attached 0004 suggestion fixes the release-critical bug #972481,
which is about to trigger the removal of reverse dependencies from
testing.
Could you please review and upload the attached suggestions,
or else approve an NMU of 0004?
Thanks.
>From 218f1f4eeff211d5fdf1fc10bb51f182333926e1 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 15 Nov 2020 09:59:56 +0100
Subject: [PATCH 1/5] Uscan v4 and Files-Excluded replace get-orig-source.sh
---
debian/README.source | 16
debian/get-orig-source.sh | 29 -
debian/watch | 8 +---
3 files changed, 5 insertions(+), 48 deletions(-)
delete mode 100644 debian/README.source
delete mode 100755 debian/get-orig-source.sh
diff --git a/debian/README.source b/debian/README.source
deleted file mode 100644
index e7d7c5e..000
--- a/debian/README.source
+++ /dev/null
@@ -1,16 +0,0 @@
-git-buildpackage
-
-
-updating package
-
-
-In order to download a repacked version of the upstream tarball execute
-
-debian/rules get-orig-source
-
-Import the generated tarball as normal with
-
-git import-orig
-
-
- -- Per Andersson Fri, 04 Apr 2014 02:05:06 +0200
diff --git a/debian/get-orig-source.sh b/debian/get-orig-source.sh
deleted file mode 100755
index c6bbf18..000
--- a/debian/get-orig-source.sh
+++ /dev/null
@@ -1,29 +0,0 @@
-#!/bin/sh
-
-set -e
-
-# bail if we already are up to date
-if ! uscan --no-symlink ; then
-echo "Already up-to-date."
-exit 0
-fi
-
-upstream_version=$(uscan --report --verbose | sed -n '/Newest version/ {s/.*site is \(.*\),.*/\1/;p}')
-upstream_tarball=../v$upstream_version.tar.gz
-version="$(uscan --report --verbose | sed -n '/Newest version/ {s/.*site is \(.*\),.*/\1/;p}')~ds1"
-repack_dir="../ruby-pygments.rb-$version"
-
-tar zxvf $upstream_tarball -C ..
-mv ../pygments.rb-* $repack_dir
-
-## remove vendor libraries
-rm -rf $repack_dir/vendor
-
-tar zcvf ../ruby-pygments.rb-$version.tar.gz $repack_dir
-
-echo "Deleting downloaded files.."
-rm -rf $repack_dir
-rm -f $upstream_tarball
-
-echo "New repackaged orig source tarball is at"
-echo "../ruby-pygments.rb-$version.tar.gz"
diff --git a/debian/watch b/debian/watch
index 2b1a980..4a84678 100644
--- a/debian/watch
+++ b/debian/watch
@@ -1,3 +1,5 @@
-version=3
-opts=dversionmangle=s/~ds\d+$// \
-https://github.com/tmm1/pygments.rb/tags .*/v(.*).tar.gz
+version=4
+
+opts="filenamemangle=s%(?:.*?)?v?(\d[\d.]*)\.tar\.gz%ruby-pygments.rb-$1.tar.gz%" \
+https://github.com/tmm1/pygments.rb/tags \
+(?:.*?/)?v?(\d[\d.]*)\.tar\.gz
--
2.20.1
>From f19fcc44542b10339b097b7e1a2e8497fd6ac69f Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 15 Nov 2020 10:05:32 +0100
Subject: [PATCH 2/5] Debhelper 13 (no changes)
---
debian/control | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/debian/control b/debian/control
index cddc095..9716c9d 100644
--- a/debian/control
+++ b/debian/control
@@ -4,7 +4,7 @@ Priority: optional
Maintainer: Debian Ruby Team
Uploaders: Per Andersson ,
Dominique Dumont
-Build-Depends: debhelper-compat (= 12),
+Build-Depends: debhelper-compat (= 13),
gem2deb (>= 1),
rake,
rake-compiler,
--
2.20.1
>From a5813241f2f76b3e8c43d763c4c76a1adebb5787 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 15 Nov 2020 10:06:31 +0100
Subject: [PATCH 3/5] Slightly shorten d/rules
---
debian/clean | 1 +
debian/rules | 7 +--
2 files changed, 2 insertions(+), 6 deletions(-)
create mode 100644 debian/clean
diff --git a/debian/clean b/debian/clean
new file mode 100644
index 000..6628c9b
--- /dev/null
+++ b/debian/clean
@@ -0,0 +1 @@
+lexers
diff --git a/debian/rules b/debian/rules
index d344b4a..a50cddd 100755
--- a/debian/rules
+++ b/debian/rules
@@ -8,13 +8,8 @@ export DH_RUBY_GEM_INSTALL_BLACKLIST_APPEND = vendor/*
%:
dh $@ --buildsystem=ruby --with ruby
-override_dh_auto_build:
+execute_before_dh_auto_build:
rake lexers
- dh_auto_build
-
-override_dh_auto_clean:
- $(RM) -f lexers
- dh_auto_clean
override_dh_compress:
dh_compress
--
2.20.1
>From fea0a6c436cdf73a3ae1dc0ad783d32b0286d4cd Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 15 Nov 2020 10:51:11 +0100
Subject: [PATCH 4/5] Make the test less dependent on the pygments version,
closes: #972481
---
debian/patches/series | 1 +
...t-less-dependency-on-pygments-version.diff | 34 +++
2 files changed, 35 insertions(+)
create mode 100644 debian/patches/test-less-dependency-on-pygments-version.diff
diff --git a/debian/patches/series b/debian/patches/series
index 7ab1c50..208d5a3 100644
--- a/debian/patches/series
+++ b/debian/patches/series
@@ -12,3
Bug#966968: efp: FTBFS: efp.a65:line 89: 8000:Illegal segment error
Package: src:efp Followup-For: Bug #966968 asm_to_a65.pl translates efp.asm (DASM format) to efp.a65 (XA65 format). The related lines are: # "org FOO" ==> "*=FOO" s/\s*org\s+/*=/; and s/\s*SEG\.U\s+\w+/.zero/; s/\s*SEG\s+/./; They rewrite ; *** DATA SEGMENT SEG.U data org $ as ; *** DATA SEGMENT .zero *=$ The '.zero' line is rejected by xa65 2.3.11. There has probably been a change in the syntax described in http://www.floodgap.com/retrotech/xa/dists/xa-2.3.11.tar.gz /doc/fileformat.txt
Bug#971018: libgnatcoll-db: FTBFS on mips(64)el with assembler message: branch out of range
Source: libgnatcoll-db Version: 20.0-1 Severity: serious Justification: fails to build from scratch on previously built architectures Hello. libgnatcoll-db/20.0-1 in experimental fails to build on mipsel and mips64el with the following messages [1]. /tmp/ccl8IVnY.s: Assembler messages: /tmp/ccl8IVnY.s:692755: Error: branch out of range /tmp/ccl8IVnY.s:693545: Error: branch out of range /tmp/ccl8IVnY.s:693669: Error: branch out of range The problem seems related to a previous bug [2] with similar symptoms [3]. xref/gnatcoll-xref.adb: In function ‘GNATCOLL.XREF’: xref/gnatcoll-xref.adb:40:1: note: variable tracking size limit exceeded with ‘-fvar-tracking-assignments’, retrying without Assembler messages: 678119: Error: branch out of range CFLAGS+=-mxgot has fixed the issue then, but does not seem sufficient anymore. Any advice would be welcome. [1] https://buildd.debian.org/status/package.php?p=libgnatcoll-db=experimental [2] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=953069 [3] https://buildd.debian.org/status/fetch.php?pkg=libgnatcoll-db=mipsel=19.2-2=1584014543=0
Bug#960872: nmudiff
Package: gtkgl2
Control: tags -1 pending
Dear maintainer,
I've prepared an NMU for gtkgl2 (versioned as 2.1.0-0.3) fixing this
bug. The diff with 2.1.0-0.1 is attached.
I have uploaded it to DELAYED/15.
Feel free to tell me if I should cancel this upload.
Regards.
diff --git a/debian/changelog b/debian/changelog
index dfe3329..cd3153d 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,20 @@
+gtkgl2 (2.1.0-0.3) unstable; urgency=medium
+
+ * Non-maintainer upload.
+ * autopkgtest: call gcc instead of cc. Closes: #960872.
+ * Use substitution variables to restrict patterns in debhelper config files.
+
+ -- Nicolas Boulenguez Mon, 18 May 2020 00:31:57 +
+
+gtkgl2 (2.1.0-0.2) unstable; urgency=medium
+
+ * Non-maintainer upload.
+ * Let autopkgtests explicitly depend on gcc.
+ * Drop obsolete libgtkgl2.0-dev.lintian-overrides.
+ * Debhelper 13.
+
+ -- Nicolas Boulenguez Sun, 03 May 2020 06:43:46 +
+
gtkgl2 (2.1.0-0.1) unstable; urgency=medium
* Non-maintainer upload.
diff --git a/debian/control b/debian/control
index b303b34..aa7d337 100644
--- a/debian/control
+++ b/debian/control
@@ -4,7 +4,7 @@ Priority: optional
Maintainer: Sam Hocevar
Standards-Version: 4.5.0
Build-Depends:
- debhelper-compat (= 12),
+ debhelper-compat (= 13),
libgl-dev,
libgtk2.0-dev,
Rules-Requires-Root: no
diff --git a/debian/libgtkgl2.0-1.install b/debian/libgtkgl2.0-1.install
index e5e5fe1..858d8e3 100644
--- a/debian/libgtkgl2.0-1.install
+++ b/debian/libgtkgl2.0-1.install
@@ -1 +1 @@
-usr/lib/*/libgtkgl*.so.*
+usr/lib/${DEB_HOST_MULTIARCH}/libgtkgl*.so.*
diff --git a/debian/libgtkgl2.0-dev.install b/debian/libgtkgl2.0-dev.install
index 7ee2227..fb6ee69 100644
--- a/debian/libgtkgl2.0-dev.install
+++ b/debian/libgtkgl2.0-dev.install
@@ -1,4 +1,4 @@
usr/include/gtkgl-*/gtkgl/*.h
-usr/lib/*/libgtkgl-*.a
-usr/lib/*/libgtkgl-*.so
-usr/lib/*/pkgconfig/gtkgl-*.pc
+usr/lib/${DEB_HOST_MULTIARCH}/libgtkgl-*.a
+usr/lib/${DEB_HOST_MULTIARCH}/libgtkgl-*.so
+usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig/gtkgl-*.pc
diff --git a/debian/libgtkgl2.0-dev.lintian-overrides b/debian/libgtkgl2.0-dev.lintian-overrides
deleted file mode 100644
index 9b22ef9..000
--- a/debian/libgtkgl2.0-dev.lintian-overrides
+++ /dev/null
@@ -1,2 +0,0 @@
-# This package Depends: libgl1 via libgl-dev.
-pkg-config-references-unknown-shared-library usr/lib/*/pkgconfig/gtkgl-2.0.pc -lGL (line 10)
diff --git a/debian/not-installed b/debian/not-installed
index 7fdf2fe..9c992f6 100644
--- a/debian/not-installed
+++ b/debian/not-installed
@@ -1 +1 @@
-usr/lib/*/libgtkgl-*.la
+usr/lib/${DEB_HOST_MULTIARCH}/libgtkgl-*.la
diff --git a/debian/tests/control b/debian/tests/control
index 213281a..74bf333 100644
--- a/debian/tests/control
+++ b/debian/tests/control
@@ -1,4 +1,4 @@
Tests: link-basic-example
-Depends: @, pkg-config, xauth, xvfb
+Depends: @, pkg-config, xauth, xvfb, gcc
Restrictions: allow-stderr
# Parsing of gtkgl headers emits deprecation warnings.
diff --git a/debian/tests/link-basic-example b/debian/tests/link-basic-example
index d2fe769..76453e5 100644
--- a/debian/tests/link-basic-example
+++ b/debian/tests/link-basic-example
@@ -6,5 +6,5 @@ sed 's/gtk_main();//' examples/simple.c > "$AUTOPKGTEST_TMP/m.c"
cd "$AUTOPKGTEST_TMP"
-cc -o m m.c `pkg-config --cflags --libs gtkgl-2.0`
+gcc -o m m.c `pkg-config --cflags --libs gtkgl-2.0`
xvfb-run ./m
Bug#954681: [gnat-9] crash caused by SOURCE_DATE_EPOCH patch
Hello. Commit 2e5e37249f09e54bb62e3ab87508e43ff709a4fb should improve the situation. It reverts the part affecting s-os_lib.adb. A GCC build produces the same checksum for it than 9.3.0-8 in libgnat-9, hopefully making gnat usable again by existing packaged libraries (I don't know how to test this on a porterbox). The remaining part, slightly adapted, passes the regression tests for this specific bug. If all goes well, please lower the severity of this bug, but keep it open as a reminder. * Unless an objection arises, the gcc-10 branch should eventually cherry-pick c1d276b0e2cb60140389aa4c23ac3addb7a77af3. * Ludovic has noticed that the initial error message CONSTRAINT_ERROR : uintp.adb:1959 overflow check failed probably reveals a bug that should be investigated by itself. No time computation is supposed to trigger a range overflow in an unrelated package.
Bug#954681: [gnat-9] crash caused by SOURCE_DATE_EPOCH patch
Package: src:gcc-9 Followup-For: Bug #954681 Control: reopen -1 Hello. The initial symptom is cured, but the fix lets debian/patches/ada-lib-info-source-date-epoch.diff modify gcc/ada/libgnat/s-os_lib.ad[bs]. This invalidates the checksums embedded in the .ali files of libgnat-9, causing a failure in all Ada compilations involving packaged dependencies. Especially, all run time tests for packaged libraries fail. Ludovic suggests to revert the fix (commit c1d276b0e2cb60140389aa4c23ac3addb7a77af3 )in the gcc-9 branch. I suggest to cherry-pick the commit in the gcc-10 branch, so that it is part of the future gnat-10 transition. Matthias, do you agree with these two actions? Then, we will search for a less intrusive work-around for this bug in gcc-9.
Bug#954681: gnat: crash caused by SOURCE_DATE_EPOCH patch
Package: src:gprbuild
Followup-For: Bug #954681
Control: retitle -1 [gnat-9] crash caused by SOURCE_DATE_EPOCH patch
Control: reassign -1 src:gcc-9 9.3.0-1
Control: affects -1 + gprbuild
Control: tags -1 + pending
Hello.
The problem is a regression in GCC from 9.3.0-1 to 9.3.0-8, visible
only when SOURCE_DATE_EPOCH is set. A reproducer is attached.
The VCS contains a fix.
https://salsa.debian.org/toolchain-team/gcc/-/commit/c1d276b0e2cb60140389aa4c23ac3addb7a77af3
echo 'Regression test for #954681'
# In gcc-*/build/testdir/:
# export LD_LIBRARY_PATH=../gcc/ada/rts
# gnatmake='../gcc/gnatmake --GCC=../gcc/xgcc --GNATBIND=../gcc/gnatbind
--GNATLINK=../gcc/gnatlink'
# Normally:
gnatmake=gnatmake
cat > gpr.ads < gpr-compilation.ads < gpr-compilation-slave.ads < gpr-compilation-slave.ali < 80|481r30 970r21
56V13*Clock{42I9} 80|596s25 610s24 999s16 1007s15
113V14*"-"{duration} 80|614s36 1014s36
223I9 Time_Rep
X 9 a-contai.ads
16K13*Containers 24e19 79|25r10 105r28 80|26r10 60r27 479r16 862r19 863r29
20I9*Count_Type 80|479r27 862r30 863r40
X 11 a-conhel.ads
84R12 Reference_Control_Type<30|43R9>
X 13 a-ciorse.ads
339R9 Node_Type
340P9 Node_Access(339R9)
X 14 a-ciorse.adb
55V13 Color{19|37E9} 22|45i19[81|57]
88V13 Left{13|340P9} 22|41i19
91V13 Parent{13|340P9[81|57]} 22|39i19[81|57]
99V13 Right{13|340P9} 22|43i19
102U14 Set_Color 22|46i19[81|57]
105U14 Set_Left 22|42i19[81|57]
108U14 Set_Parent 22|40i19[81|57]
111U14 Set_Right 22|44i19[81|57]
X 15 a-coinve.ads
57R9*Vector<30|43R9> 79|65r33[91|39] 112r50[81|41] 113r50[81|41] 114r50[81|41]
. 80|402r50[81|41] 403r50[81|41] 404r50[81|41] 484r37[81|41] 488r50[81|41]
. 489r50[81|41] 490r50[81|41] 497r37[81|41] 631r33[91|39]
100V13*Length{9|20I9} 80|586s28[81|41] 587s36[81|41]
234U14*Append 80|507s18[81|41]
X 17 a-coorse.ads
47k24*Ordered_Sets 453e32 80|26w21 60r38
54R9*Set<30|43R9> 80|135r22[60]
61R9*Cursor 80|798r38[60] 836r38[60] 931r38[60] 942r38[60]
64V13*Has_Element{boolean} 80|838s21[60]
79V13*Length{9|20I9} 80|818s31[60] 863s59[60]
83U14*Clear 80|809s15[60]
85V13*Element{80|44R9} 80|839s28[60] 918s37[60]
87U14*Replace_Element 80|921s21[60]
119U14*Insert 80|907s15[60]
191V13*Find{61R9[80|60]} 80|798s53[60] 836s53[60] 931s53[60] 942s53[60]
219V13*Iterate{32|32C9[70[80|60]]} 80|916s24[60]
325R9 Node_Type
326P9 Node_Access(325R9)
X 18 a-coorse.adb
58V13 Color{19|37E9} 22|45i19
61V13 Left{17|326P9[80|60]} 22|41i19[80|60]
64V13 Parent{17|326P9[80|60]} 22|39i19[80|60]
67V13 Right{17|326P9[80|60]} 22|43i19[80|60]
70U14 Set_Color 22|46i19[80|60]
73U14 Set_Left 22|42i19
76U14 Set_Right 22|44i19
79U14 Set_Parent 22|40i19
X 26 a-convec.ads
73k24*Vectors 824e27 79|25w21 105r39
88R9*Vector<30|43R9> 79|107r27[105]
431U14*Append 80|383s25[79|105]
X 28 a-direct.ads
76K13*Directories 502e20 80|27w10 27r43
82V13*Current_Directory{string} 80|636s32
251V13*Containing_Directory{string} 80|518s10 761s21 762s16
310V13*Exists{boolean} 80|549s38 674s22 701s22 721s28
X 29 a-except.ads
54K13*Exceptions 349e19 80|28w10 28r43
104V13*Exception_Information{string} 80|783s47 1143s20
X 32 a-iteint.ads
24h9 Forward_Iterator
32h9 Reversible_Iterator<24R9[13|70]>
32C9*Reversible_Iterator<32R9[17|70[80|60]]>
32h9*Reversible_Iterator<24R9[17|70[80|60]]>
X 33 a-numeri.ads
16K13*Numerics 32e17 80|29r10 29r43
X 34 a-nuflra.ads
41K22*Float_Random 74e30 80|29w19 29r52
47R9*Generator<129|63R9> 80|97r21
49F12*Uniformly_Distributed{float}
51V13*Random{49F12}<129|71p13> 80|865s53
X 38 a-string.ads
16K13*Strings 35e16 79|26r10 37r12 80|30r10 30r43 31r10 31r43 347r16
X 39 a-strfix.ads
18K21*Fixed 251e22 80|30w18 30r51
36V13*Index{natural} 80|332s34 658s36
X 40 a-strmap.ads
38K21*Maps 411e21 80|31r18 347r24 348r16 349r16
46A9*Character_Set<137A9>
86V13*Is_Subset{boolean} 80|347s29
98V13*To_Set{46A9} 80|348s21
137A9 Character_Set_Internal(boolean)
X 41 a-stmaco.ads
38K26*Constants 915e31 80|31w23 349r21
48a4*Decimal_Digit_Set{40|46A9} 80|349r31
X 43 a-strunb.ads
76K21*Unbounded 491e26 79|26w18 37r20
79R9*Unbounded_String<30|43R9> 474e14 79|91r19 98r14 80|50r23 333r14 408r13
82r4*Null_Unbounded_String{79R9} 80|247r24 412r20
84V13*Length{natural} 80|440s10
94V13*To_Unbounded_String{79R9} 80|291s27 339s18 342s18 517s19 541s46 543s46
100V13*To_String{string} 80|190s44 194s53 253s28 265s18 272s15 296s21 304s26
. 421s40 430s35 433s37 453s34 468s40 545s40 549s46 550s59 553s49 560s53 639s32
. 1098s27 1108s40
111U14*Append 80|413s13 415s10
166V14*"="{boolean} 80|247s22 412s17
X 46 a-textio.ads
49K13*Text_IO 471e16 80|32w10 32r43 697r31 709r22 709r42 710r22 711r22 712r22
52P9*File_Type 80|697r39
53n23*In_File{53E9} 80|709r50
92U14*Open 80|709s30
98U14*Close 80|712s30
168U14*Skip_Line 80|710s30
242U14*Put 80|430s10 431s10 613s10 614s10 1013s10 1014s10 1098s22
244U14*Get_Line 80|711s30
263U14*Put_Line 80|189s19 194s16 198s16 248s10 263s13 270s10 294s16 300s13
. 309s13 380s13 432s10 433s10 441s10 446s10 552s34 559s34 589s10 615s10 681s22
. 683s22 727s28 729s28 735s22 737s22 783s10 1015s10 1133s16 1143s10
X 49 gnat.ads
Bug#936869: libgnatcoll-bindings: Python2 removal in sid/bullseye
Package: src:libgnatcoll-bindings Followup-For: Bug #936869 User: debian-pyt...@lists.debian.org Usertags: py2keep Popcon for gnat-gps is just above the 300 limit. The main reason to package libgnatcoll-bindings is that gnat-gps depends on the python module. libgnatcoll-bindings is mostly ready for python3, but not gnat-gps. For now, libgnatcoll-bindings is stuck with Python2. Upstream of both packages intends to update the python support for the middlef 2021 if all goes well. Libgnatcoll-bindings and gnat-gps will most probably have to skip next Debian release.
Bug#936869: libgnatcoll-bindings: Python2 removal in sid/bullseye
Package: src:libgnatcoll-bindings Followup-For: Bug #936869 Control: tags -1 + pending Hello. An upload to unstable is planned as part of the gnat-9 transition. It will include explicit python2 dependencies.
Bug#945197: nmu libgnatcoll...
Hello. The fix in 2018-7 was incomplete, even with some NMUs and rebuilds. Please ignore this request. Sorry for the noise.
Bug#945197: fixed in gprbuild 2018-7
> > Rename libgpr.(a|so) to libgnatprj because the grpc source owns them. > your reverse dependencies [...] don't find the now removed library files This upload was an error. I am going to revert the changes and let libgpr3-dev conflict with libgprc-dev. The libgpr->libgnatprj renaming is implemented in experimental. All dependencies will be updated as part of the gcc-9 transition. Thanks for your explanations.
Bug#945197: libgrpc-dev,libgpr3-dev: both ship libgpr.a, libgpr.so
Hello. I have contacted the upstream for gpbruild (AdaCore) months ago, without answer for now. Google (for grpc) will probably also have more urgent concerns that avoiding a name clash with an unrelated library. I hope that the issue has been reported there too, so that they avoid third-letter global identifiers in the future :-). In Debian, libgpr has been introduced by: grpc on 2015-08-16 gprbuild on 2016-08-11 Unless someone suggests a better fix, I intend to rename gprbuild/libgpr to gprbuild/libgnatprj.
Bug#942442: Fwd: Bug#942442: libgnatvsn ftbfs in 9.2.1-10
> libgnatvsn ftbfs in 9.2.1-10 Sorry for the mess. After a successful build, it seems that I have incorrectly played the autoreconf/quilt refresh/scp/git commit music. A build currently tests the following fix. --- a/debian/patches/ada-libgnatvsn.diff +++ b/debian/patches/ada-libgnatvsn.diff @@ -437,7 +437,7 @@ } --- /dev/null +++ b/src/libgnatvsn/Makefile.in -@@ -0,0 +1,899 @@ +@@ -0,0 +1,938 @@ +# Makefile.in generated by automake 1.16.1 from Makefile.am. +# @configure_input@ +
Bug#932452: adacontrol: FTBFS on i386 mips because of tests
Source: adacontrol Version: 1.21r3-2 Severity: serious Justification: fails to build from scratch on i386 mips Here are the relevant lines from build logs. On i386 mips hppa powerpc: => t_type_usage.txtFAILED<= ... -- Passed= 85, Failed= 1 -- --- Total run time: 1mn 20s. make[1]: *** [debian/rules:58: override_dh_auto_test-arch] Error 1 On sparc64: => tfw_stress.txt FAILED - Passed= 85, Failed= 1 -- On s390x: -- Passed= 86, Failed= 0 -- --- make[1]: *** [debian/rules:58: override_dh_auto_test-arch] Error 1
Bug#927943: libxmlada: FTBFS with unicode-data >= 12.0.0
Package: src:libxmlada
Followup-For: Bug #927943
The attached patch builds libxmlada/18-3 against unicode-data/12.1.0.
According to a diff of generated sources, no existing symbol is
changed or removed, so it should be safe to keep the same SO version
for the binary library package.
I hope that no new ALI version is necessary for the -dev package. Each
recursive reverse dependency would then require trivial but manual
changes.
gpr *
dh-ada-library
templates-parser *
gnat-gps
gnatcoll *
gnatcoll-bindings *
gnatcoll-db *
asis *
aws *
adacontrol
adabrowse
log4ada *
Worst, the libraries (with a star) would require a passage through
NEW. Such a transition is impossible so far in the freeze, so gpr
FTBFS in testing. Summary: buster would not support the Ada language.
I see no better option than uploading the fix with the same ALI,
and hope that all goes well.
*
https://codesearch.debian.net/search?q=path%3A.*%5C.ad%5Bbs%5D+with%5C+Unicode%5C.Names%5C.=3
seems to show that the generated packages are only directly referenced
from libxmlada itself.
* The diff_alis.py script (once posted in debian-ada) says that no
existing .ali checksum is modified between libxmlada/18-2 and 18-3
(I fail to understand why most unicode-names-*.ads have no .ali file
at all, but it is quite convenient right now).
* If some reverse dependencies FTBFS and must be removed, what do we
loose compared to the current situation?
Of course, any better idea is welcome.
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,9 @@
+libxmlada (18-4) unstable; urgency=medium
+
+ * Update to unicode version 12. Closes: #927943.
+
+ -- Nicolas Boulenguez Thu, 02 May 2019 11:47:15 +0200
+
libxmlada (18-3) unstable; urgency=medium
* Update to unicode version 11, really closing #903380.
--- a/debian/control
+++ b/debian/control
@@ -7,7 +7,7 @@
Build-Depends: debhelper (>= 11),
gnat, gnat-8,
# Note: the rules script parses this line to extract the gnat version.
- unicode-data (>= 11~), unicode-data (<< 12~),
+ unicode-data (>= 12~), unicode-data (<< 13~),
# Used by debian/rules to generate unicode-names-*.ads.
# When a new unicode standard is released, we most probably have to
# increase the ALI version of xmlada-unicode and all reverse dependencies.
--- a/debian/patches/generate-unicode-9-names.diff
+++ b/debian/patches/generate-unicode-9-names.diff
@@ -14,7 +14,7 @@
end Translators.Alias;
--- a/unicode/importer/translators-block.adb
+++ b/unicode/importer/translators-block.adb
-@@ -164,6 +164,10 @@
+@@ -164,6 +164,15 @@
"Phags_Pa");
Add ("CJK Compatibility Forms",
"Cjk_Compatibility_Forms");
@@ -22,6 +22,11 @@
+ -- Unicode 9.
+ Add ("Ideographic Symbols and Punctuation",
+ "Ideograph_Symb_Punct");
++ -- Unicode 12.
++ Add ("Egyptian Hieroglyph Format Controls",
++ "Egypt_Hieroglyph_Fmt_Ctrl");
++ Add ("Symbols and Pictographs Extended-A",
++ "Symbols_Pictographs_Ext_A");
end Set_Exceptions;
end Translators.Block;
Bug#927229: exadrums: FTBFS on armel mips mipsel because of missing -latomic
Package: exadrums Followup-For: Bug #927229 Control: reassign -1 libexadrums Control: retitle -1 libexadrums requires libatomic on some archiectures Control: affects -1 exadrums Control: tag -1 + patch Hello. The exadrums executable directly uses neither threads nor atomic variables. The problem is a missing dependency in the libexadrums library. https://buildd.debian.org/status/fetch.php?pkg=libexadrums=armel=0.3.0-1=1555353342=0 A patch is attached. Once a fixed libexadrums is available in the archive, it should be sufficient to rebuild exadrums on the affected architectures. I may sponsor the upload if/when you agree on the solution. By the way… In order to have such issues reported sooner, I suggest adding -Wl,--no-undefined to the linker flags. The -pthread "adds architecture-dependent compiler and linker options required when LIBS+=-lpthread is intended". As far as I understand, no one should ever use -lpthread only. https://stackoverflow.com/questions/2127797/significance-of-pthread-flag-when-compiling and especially http://www.hpl.hp.com/techreports/2004/HPL-2004-209.pdf The reason why libexadrums requires *both* libatomic and libpthread should probably be documented somewhere. diff --git a/debian/changelog b/debian/changelog index d4cc7aa..0948542 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,9 @@ +exadrums (0.3.0-2) experimental; urgency=medium + + * Add -pthread -latomic --no-undefined build flags. Closes: #927229. + + -- Jeremy Oden Fri, 19 Apr 2019 12:44:53 +0200 + libexadrums (0.3.0-1) experimental; urgency=medium * Initial Debian packaging. Closes: #919327. diff --git a/debian/patches/libatomic.diff b/debian/patches/libatomic.diff new file mode 100644 index 000..014e123 --- /dev/null +++ b/debian/patches/libatomic.diff @@ -0,0 +1,35 @@ +Description: explicitly add -latomic to linker flags. + The upstream author is also the Debian maintainer, so there is no + need to forward the patch further. +Bug-Debian: https://bugs.debian.org/927229 +Forwarded: not-needed +Author: Nicolas Boulenguez + +--- a/Makefile.am b/Makefile.am +@@ -23,16 +23,23 @@ + + # Good habits. + AM_CXXFLAGS = -Wall +-AM_LDFLAGS = -Wl,--as-needed ++AM_LDFLAGS = -Wl,--as-needed -Wl,--no-undefined + + # Delegate library handling to libtool. + lib_LTLIBRARIES = libexadrums.la + + libexadrums_la_CXXFLAGS = $(AM_CXXFLAGS) \ ++ -pthread \ + -std=c++17 $(alsa_CFLAGS) $(tinyxml2_CFLAGS) + libexadrums_la_LDFLAGS = $(AM_LDFLAGS) \ ++ -pthread \ + -version-info $(exadrums_libtool_version_info) +-libexadrums_la_LIBADD = $(alsa_LIBS) $(tinyxml2_LIBS) -lpthread -lstdc++fs ++# The -pthread compiler and linker option links with libpthread, but ++# also selects various architecture-dependent settings. ++libexadrums_la_LIBADD = $(alsa_LIBS) $(tinyxml2_LIBS) -latomic -lstdc++fs ++# The standard C++ library does not include libatomic on some ++# architectures (armel mips mipsel). It does not hurt to always list ++# it here, it will only be effective when used thanks to --as-needed. + + # Distribute and embed into the library. + libexadrums_la_SOURCES = \ diff --git a/debian/patches/series b/debian/patches/series new file mode 100644 index 000..2829210 --- /dev/null +++ b/debian/patches/series @@ -0,0 +1 @@ +libatomic.diff
Bug#927229: exadrums: FTBFS on armel mips mipsel because of missing -latomic
Package: exadrums Version: 0.3.0-1 Severity: serious Justification: failure to build from scratch on release architecture Hello. The package source fails to build from scratch on armel mips mipsel (as well as other non released architectures) https://buildd.debian.org/status/package.php?p=exadrums=experimental with the following message /usr/bin/ld: /usr/lib/gcc/.../libexadrums.so: undefined reference to `__atomic_store_8' /usr/bin/ld: /usr/lib/gcc/.../libexadrums.so: undefined reference to `__atomic_load_8' /usr/bin/ld: /usr/lib/gcc/.../libexadrums.so: undefined reference to `__atomic_exchange_8' The solution is most probably to add -pthread to both CXXFLAGS and LDFLAGS. In LDFLAGS, it should come after --as-needed so the dependency is only effective on architectures actually using these 3 symbols.
Bug#924835: adacontrol: FTBFS: test tfw_gpr.txt failed
Package: src:adacontrol Followup-For: Bug #924835 Control: severity 924835 minor 1.20r7-2 disables all build time tests.
Bug#924835: adacontrol: FTBFS: test tfw_gpr.txt failed
Package: src:adacontrol Followup-For: Bug #924835 Hello. I fail to reproduce the attached log with adacontrol/1.20r7-1 in a local amd64 buster chroot. In my chroot and during reproducible builds, the failure of the tfw_gpr test is reported but not critical thanks to the fix_tfw_gpr.diff Debian patch. However, it breaks the build on several non-released architectures with 1.20r7-1 in unstable. A proper fix is introduced by 1.20r9-1 in experimental. The t_style test succeeds in my chroot and during reproducible builds. It fails on several non-released architectures in both unstable and experimental. Unless someone manages to reproduce both issues (probably in a porterbox) and upload a targeted fix within 5 days, the package will be removed from buster without possible return. This seems hazardous, so for now I suggest to simply disable all tests during the build. Any better idea is welcome.
Bug#920337: python3-igraph: ships header in /usr/include/python3.7
Hello.
The attached patch works around this bug in python-igraph.
The issue should probably be fixed globally by adding 'm'
in package python3-stdlib-extensions
in filedebian/patches/3.7/distutils-install-layout.diff
in line'headers': '$base/include/python$py_version_short/$dist_name'
but I am not an expert in Python packaging, so I would like your
opinion before opening a bug against
python3-stdlib-extensions.
Ideally, the bug number would be added to a comment in debian/rules so
that we know when the work-around can be removed.
--- a/debian/compat
+++ /dev/null
@@ -1,1 +0,0 @@
-11
--- a/debian/control
+++ b/debian/control
@@ -4,7 +4,7 @@
Maintainer: Debian Python Modules Team
Uploaders: TANIGUCHI Takaki ,
Hugo Lefeuvre
-Build-Depends: debhelper (>= 11),
+Build-Depends: debhelper-compat (= 12),
dh-python,
libigraph-dev,
pkg-config,
@@ -17,6 +17,7 @@ Build-Depends: debhelper (>= 11),
python3-setuptools,
python3-texttable
Standards-Version: 4.3.0
+Rules-Requires-Root: no
Testsuite: autopkgtest-pkg-python
Homepage: http://igraph.org/python/
Vcs-Git: https://salsa.debian.org/python-team/modules/python-igraph.git
--- a/debian/rules
+++ b/debian/rules
@@ -3,11 +3,19 @@
export DEB_BUILD_MAINT_OPTIONS = hardening=+all
export PYBUILD_NAME=igraph
+# Work around #920337. The proper fix probably adds 'm'
+# in package python3-stdlib-extensions
+# in filedebian/patches/3.7/distutils-install-layout.diff
+# in line'headers': '$base/include/python$py_version_short/$dist_name'
+export PYBUILD_INSTALL_ARGS_python3 := \
+ --install-headers=/usr/include/python{version}m
+
%:
# Ensure that the embedded copy is never used.
rm -f vendor/texttable.py
dh $@ --with python2,python3 --buildsystem=pybuild
+# Disable the scripts= option in setup.cfg. See #664443.
override_dh_auto_install:
dh_auto_install
rm -f $(CURDIR)/debian/python-igraph/usr/bin/igraph
--- a/debian/watch
+++ b/debian/watch
@@ -1,3 +1,2 @@
-version=3
-https://pypi.debian.net/python-igraph/python-igraph-(.*)\.tar\.gz
-
+version=4
+https://pypi.debian.net/@PACKAGE@/@PACKAGE@@ANY_VERSION@@ARCHIVE_EXT@
Bug#919219: abcm2ps: diff for NMU version 8.14.2-0.2
Dear maintainer,
I've prepared an NMU for abcm2ps (versioned as 8.14.2-0.2) and
uploaded it to DELAYED/10. Please feel free to tell me if I
should delay it longer.
It mostly fixes my previous the build with an explicit debhelper build
system (#919219).
Regards.
diff -Nru abcm2ps-8.14.2/debian/changelog abcm2ps-8.14.2/debian/changelog
--- abcm2ps-8.14.2/debian/changelog 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/changelog 2019-01-27 18:50:50.0 +0100
@@ -1,3 +1,13 @@
+abcm2ps (8.14.2-0.2) unstable; urgency=medium
+
+ * Non-maintainer upload.
+ * Drop obsolete README.Debian. Closes: #919514.
+ * Debhelper 12.
+ * Cherry-pick bad-ps-when-centered-or-right-aligned.diff from upstream CVS.
+ * Fix the build with an explicit debhelper build system. Closes: #919219.
+
+ -- Nicolas Boulenguez Sun, 27 Jan 2019 18:50:50 +0100
+
abcm2ps (8.14.2-0.1) unstable; urgency=medium
* Non-maintainer upload.
diff -Nru abcm2ps-8.14.2/debian/compat abcm2ps-8.14.2/debian/compat
--- abcm2ps-8.14.2/debian/compat 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/compat 1970-01-01 01:00:00.0 +0100
@@ -1 +0,0 @@
-11
diff -Nru abcm2ps-8.14.2/debian/control abcm2ps-8.14.2/debian/control
--- abcm2ps-8.14.2/debian/control 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/control 2019-01-27 18:39:38.0 +0100
@@ -3,7 +3,7 @@
Priority: optional
Maintainer: Anselm Lingnau
Build-Depends:
- debhelper (>= 11),
+ debhelper-compat (= 12),
libfreetype6-dev,
libpango1.0-dev,
pkg-config,
diff -Nru abcm2ps-8.14.2/debian/patches/bad-ps-when-centered-or-right-aligned.diff abcm2ps-8.14.2/debian/patches/bad-ps-when-centered-or-right-aligned.diff
--- abcm2ps-8.14.2/debian/patches/bad-ps-when-centered-or-right-aligned.diff 1970-01-01 01:00:00.0 +0100
+++ abcm2ps-8.14.2/debian/patches/bad-ps-when-centered-or-right-aligned.diff 2019-01-27 18:45:27.0 +0100
@@ -0,0 +1,16 @@
+Description: fix: bad PS output when centered or right aligned text with pango
+ Reported by Hudson Lacerda.
+Author: Jean-Francois Moine
+Origin: upstream https://github.com/leesavide/abcm2ps/commit/2eb380926b02edde530d4e1bfbe54d9f9ddf94fc
+
+--- a/subs.c
b/subs.c
+@@ -566,7 +566,7 @@
+ wi /= 2;
+ // w = (float) wi / PG_SCALE;
+ w = (float) wi / PANGO_SCALE;
+- a2b("-%.1f 0 RM ", w);
++ a2b(" -%.1f 0 RM", w);
+ break;
+ }
+ pg_line_output(line);
diff -Nru abcm2ps-8.14.2/debian/patches/series abcm2ps-8.14.2/debian/patches/series
--- abcm2ps-8.14.2/debian/patches/series 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/patches/series 2019-01-27 18:48:44.0 +0100
@@ -1 +1,2 @@
fix-loss-of-sep.diff
+bad-ps-when-centered-or-right-aligned.diff
diff -Nru abcm2ps-8.14.2/debian/README.Debian abcm2ps-8.14.2/debian/README.Debian
--- abcm2ps-8.14.2/debian/README.Debian 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/README.Debian 1970-01-01 01:00:00.0 +0100
@@ -1,9 +0,0 @@
-abcm2ps for Debian
---
-
-This is a Debianization of Jef Moine's abcm2ps package. It differs
-from the original in the following respects:
-
- - abcm2ps assumes A4 paper by default.
-
- -- Anselm Lingnau , Mon Jan 5 03:02:59 2004
diff -Nru abcm2ps-8.14.2/debian/rules abcm2ps-8.14.2/debian/rules
--- abcm2ps-8.14.2/debian/rules 2018-12-29 14:56:32.0 +0100
+++ abcm2ps-8.14.2/debian/rules 2019-01-27 18:39:38.0 +0100
@@ -4,4 +4,4 @@
export DEB_LDFLAGS_MAINT_APPEND := -Wl,--as-needed
%:
- dh $@
+ dh $@ --buildsystem=autoconf
Bug#919235: abcm2ps: FTBFS because debhelper uses ninja build system instead of autoconf
Package: abcm2ps Version: 8.14.2-0.1 Severity: serious Tags: patch Justification: failure to build from scratch Hello. My NMU fails to build because debhelper uses ./build.ninja instead of ./configure and ./Makefile.in. This is quite surprising, and I cannot reproduce the issue in my chroot. However, the attached patch selects an explicit build system and most probably fixes the issue. It upgrades to debhelper 12 as well. Do you want to upload the fix? May I prepare another NMU? Shall I set a shorter delay? --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,11 @@ +abcm2ps (8.14.2-0.2) unstable; urgency=medium + + * Non-maintainer upload. + * Debhelper 12. + * Fix the build with an explicit debhelper build system. + + -- Nicolas Boulenguez Sun, 13 Jan 2019 18:10:08 +0100 + abcm2ps (8.14.2-0.1) unstable; urgency=medium * Non-maintainer upload. --- a/debian/compat +++ /dev/null @@ -1,1 +0,0 @@ -11 --- a/debian/control +++ b/debian/control @@ -3,7 +3,7 @@ Section: text Priority: optional Maintainer: Anselm Lingnau Build-Depends: - debhelper (>= 11), + debhelper-compat (= 12), libfreetype6-dev, libpango1.0-dev, pkg-config, --- a/debian/rules +++ b/debian/rules @@ -4,4 +4,4 @@ export DEB_BUILD_MAINT_OPTIONS := hardening=+all export DEB_LDFLAGS_MAINT_APPEND := -Wl,--as-needed %: - dh $@ + dh $@ --buildsystem=autoconf
Bug#912410: python-igraph's tests fail with Python 3.7
Hello.
The attached patch removes the failing tests.
Applying it would reduce the severity of this bug to normal and allow
the package to migrate to testing.
Some unrelated suggestions are also attached.
>From abb9ca492ba67702ce0068000ec81cd1124b50c0 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 30 Dec 2018 14:53:09 +0100
Subject: [PATCH 1/5] Update debhelper and standards version.
Debhelper 11.
Standards-Version: 4.3.0.
Rules-Requires-Root: no.
---
debian/compat | 2 +-
debian/control | 6 --
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/debian/compat b/debian/compat
index f599e28..b4de394 100644
--- a/debian/compat
+++ b/debian/compat
@@ -1 +1 @@
-10
+11
diff --git a/debian/control b/debian/control
index 0e48897..7fbe50e 100644
--- a/debian/control
+++ b/debian/control
@@ -4,7 +4,7 @@ Priority: optional
Maintainer: Debian Python Modules Team
Uploaders: TANIGUCHI Takaki ,
Hugo Lefeuvre
-Build-Depends: debhelper (>= 10),
+Build-Depends: debhelper (>= 11),
dh-python,
python-all-dev,
python-setuptools,
@@ -12,7 +12,9 @@ Build-Depends: debhelper (>= 10),
python3-setuptools,
libigraph-dev,
pkg-config
-Standards-Version: 4.0.0
+Standards-Version: 4.3.0
+Rules-Requires-Root: no
+Testsuite: autopkgtest-pkg-python
Homepage: http://igraph.org/python/
Vcs-Git: https://salsa.debian.org/python-team/modules/python-igraph.git
Vcs-Browser: https://salsa.debian.org/python-team/modules/python-igraph
--
2.19.2
>From 4381d9aaefbc7e7f864eac8d1795456b605f5015 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 30 Dec 2018 14:56:01 +0100
Subject: [PATCH 2/5] Build-Depend numpy in order to activate some tests.
---
debian/control | 2 ++
1 file changed, 2 insertions(+)
diff --git a/debian/control b/debian/control
index 7fbe50e..c89097f 100644
--- a/debian/control
+++ b/debian/control
@@ -7,8 +7,10 @@ Uploaders: TANIGUCHI Takaki ,
Build-Depends: debhelper (>= 11),
dh-python,
python-all-dev,
+ python-numpy,
python-setuptools,
python3-all-dev,
+ python3-numpy,
python3-setuptools,
libigraph-dev,
pkg-config
--
2.19.2
>From f355177571ea92a045ac7b60f16190791daf34fb Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 30 Dec 2018 14:57:50 +0100
Subject: [PATCH 3/5] Use packaged texttable.py instead of the embedded copy.
---
debian/control | 8 ++--
debian/rules | 2 ++
2 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/debian/control b/debian/control
index c89097f..37bb6f3 100644
--- a/debian/control
+++ b/debian/control
@@ -9,9 +9,11 @@ Build-Depends: debhelper (>= 11),
python-all-dev,
python-numpy,
python-setuptools,
+ python-texttable,
python3-all-dev,
python3-numpy,
python3-setuptools,
+ python3-texttable,
libigraph-dev,
pkg-config
Standards-Version: 4.3.0
@@ -23,7 +25,8 @@ Vcs-Browser: https://salsa.debian.org/python-team/modules/python-igraph
Package: python-igraph
Architecture: any
-Depends: ${python:Depends}, ${shlibs:Depends}, ${misc:Depends}
+Depends: ${python:Depends}, ${shlibs:Depends}, ${misc:Depends},
+ python-texttable
Provides: ${python:Provides}
Description: High performance graph data structures and algorithms (Python 2)
Python interface to the igraph high performance graph library, primarily
@@ -37,7 +40,8 @@ Description: High performance graph data structures and algorithms (Python 2)
Package: python3-igraph
Architecture: any
-Depends: ${python3:Depends}, ${shlibs:Depends}, ${misc:Depends}
+Depends: ${python3:Depends}, ${shlibs:Depends}, ${misc:Depends},
+ python3-texttable
Provides: ${python3:Provides}
Description: High performance graph data structures and algorithms (Python 3)
Python interface to the igraph high performance graph library, primarily
diff --git a/debian/rules b/debian/rules
index a6a7808..be010ba 100755
--- a/debian/rules
+++ b/debian/rules
@@ -4,6 +4,8 @@ export DEB_BUILD_MAINT_OPTIONS = hardening=+all
export PYBUILD_NAME=igraph
%:
+# Ensure that the embedded copy is never used.
+ rm -f vendor/texttable.py
dh $@ --with python2,python3 --buildsystem=pybuild
override_dh_auto_install:
--
2.19.2
>From 387f843328f905b8bf0262b05c9f6909a67d4fb4 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 30 Dec 2018 14:58:46 +0100
Subject: [PATCH 4/5] Link with --as-needed.
There remains to fix the order of options in the gcc -shared command
line to prevent the unneeded dependency on pthread.
---
debian/rules | 3 +++
1 file changed, 3 insertions(+)
diff --git a/debian/rules b/debian/rules
index be010ba..56e6325 100755
--- a/debian/rules
+++ b/debian/rules
@@ -1,6 +1,9 @@
#!/usr/bin/make -f
export DEB_BUILD_MAINT_OPTIONS = hardening=+all
+# TODO: find a way to move this option before -pthread, so that t
Bug#913371: gnat-gps fails to build on 32-bit kernels, address space exhaustion.
Package: src:gnat-gps Followup-For: Bug #913371 gnat-gps/18 now fails to build on armhf, while 18-3 was succeeding on the same architecture. After investigation, the only successful builds of gnat-gps/18 on 32 bits architectures so far (armhf i386 powerpc) were registered with a 32 bits chroot on a 64 bits machine. I suggest the following architecture list, with only 64 bits architectures, but any better idea is welcome. amd64 arm64 mips64el ppc64el s390x alpha ia64 kfreebsd-amd64 ppc64 riscv64 sparc64
Bug#914680: libplplotada3: missing Breaks+Replaces: libplplotada2
Package: libplplotada3
Version: 5.13.0+dfsg-10
Followup-For: Bug #914680
The short attachment contains untested suggestions
* for this bug
* for #821278 (Ada project)
* to set build flags (maybe fixing the empty -dbgsym packages)
* for a cosmetic change in debian/watch
>From 809d9fca94489f0845d8c5c053989150fc1b8130 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 2 Dec 2018 23:39:09 +0100
Subject: [PATCH 1/4] Improve readability of watch file thanks to version 4 of
uscan format.
---
debian/watch | 10 +++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/debian/watch b/debian/watch
index 0b81469a..a6be85a3 100644
--- a/debian/watch
+++ b/debian/watch
@@ -1,3 +1,7 @@
-version=3
-opts="uversionmangle=s/-/~/,dversionmangle=s/\+dfsg$//,repacksuffix=+dfsg" \
- http://sf.net/plplot/plplot-(.*)\.(?:zip|tgz|tbz|txz|(?:tar\.(?:gz|bz2|xz)))
+version=4
+
+opts="\
+ uversionmangle=s/-/~/,\
+ dversionmangle=s/\+dfsg$//,\
+ repacksuffix=+dfsg" \
+http://sf.net/@PACKAGE@/@PACKAGE@@ANY_VERSION@@ARCHIVE_EXT@
--
2.19.2
>From 961efce2cecfb5b3acae4086d2bd842f602b1977 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 2 Dec 2018 23:39:43 +0100
Subject: [PATCH 2/4] Complete the link step of the Ada project. Closes:
#821278.
Copy bindings/ada/CMakeLists.txt, which only passes -lplplot to
pkg-config.
---
debian/plplotada.gpr.in | 6 +-
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/debian/plplotada.gpr.in b/debian/plplotada.gpr.in
index 9d0bae03..e3877141 100644
--- a/debian/plplotada.gpr.in
+++ b/debian/plplotada.gpr.in
@@ -5,11 +5,7 @@ library project Plplotada is
for Library_Ali_Dir use "/usr/lib/@DEB_HOST_MULTIARCH@/adalib/plplotada";
for Library_Dir use "/usr/lib/@DEB_HOST_MULTIARCH@");
for Externally_Built use True;
-TODO:
- -- If options must be added to the command line after
- -- -L$(Library_Dir) -l$(Library_Name) when linking an executable:
package Linker is
- for Linker_Options use ("-lplplot", "-lor", "-lsomething", "-lsimilar");
+ for Linker_Options use ("-lplplot");
end Linker;
- -- Else remove all from TODO: to this line.
end Plplotada;
--
2.19.2
>From 28451d48642cefeac83f5704f4430435696d8e71 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 2 Dec 2018 23:41:17 +0100
Subject: [PATCH 3/4] Delegate selection of default build flags to
dpkg-buildflags.
Also, enable hardening flags. This setting can cause problems,
dpkg-buildflags(1) describes howto refine it.
---
debian/rules | 18 --
1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/debian/rules b/debian/rules
index 6580e849..b9e2274c 100755
--- a/debian/rules
+++ b/debian/rules
@@ -6,15 +6,21 @@
# Paths for Octave
OCTDIR = $(shell octave-config --print LOCALAPIOCTFILEDIR)
-# Note cmake ignores CPPFLAGS so add them to CFLAGS and CXXFLAGS as a
-# work around
-CPPFLAGS += $(shell mkoctfile -p INCFLAGS)
-CFLAGS += $(CPPFLAGS) -fvisibility=hidden
+DEB_BUILD_MAINT_OPTIONS := hardening=+all
+DPKG_EXPORT_BUILDFLAGS:= 1
+DEB_CFLAGS_MAINT_APPEND := -fvisibility=hidden
+DEB_FFLAGS_MAINT_APPEND := -fvisibility=hidden
# Don't add -fvisibility=hidden to CXXFLAGS for now as this breaks the
# octave bindings.
+DEB_CPPFLAGS_MAINT_APPEND := $(shell mkoctfile -p INCFLAGS)
+DEB_LDFLAGS_MAINT_APPEND := -Wl,--as-needed
+include /usr/share/dpkg/default.mk
+
+# Note cmake ignores CPPFLAGS so add them to CFLAGS and CXXFLAGS as a
+# work around, see #653916.
+CFLAGS += $(CPPFLAGS)
CXXFLAGS += $(CPPFLAGS)
-FFLAGS += -fvisibility=hidden
-LDFLAGS += -Wl,--as-needed
+
HOME=$(shell mktemp -d)
export verbose_test = on
--
2.19.2
>From 24225167500a55815cf87b56db53860221637b46 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Sun, 2 Dec 2018 23:42:53 +0100
Subject: [PATCH 4/4] Set the shared object version of the Ada library. Closes:
#914680.
The policy requires that the shared object version of the Ada library
matches the name of the binary package installing it, preventing such
bugs.
---
debian/rules | 8
1 file changed, 8 insertions(+)
diff --git a/debian/rules b/debian/rules
index b9e2274c..c2eb8053 100755
--- a/debian/rules
+++ b/debian/rules
@@ -34,6 +34,14 @@ CONFIGURE_OPTIONS = \
-DBUILD_DOC=ON \
-DCMAKE_INSTALL_LIBDIR=/usr/lib/$(DEB_HOST_MULTIARCH)
+# The soversion used to build and install the Ada shared library must
+# match the end of the library package name.
+ada_lib_pkg_vsn := $(shell sed -n \
+ '/^Package: libplplotada\([0-9.]\+\)$$/{s//\1/;p;q}' debian/control)
+ada_soname := libplplotada.so.$(ada_soversion)
+CONFIGURE_OPTIONS += \
+ '-DCMAKE_SHARED_LIBRARY_SONAME_Ada_FLAG=-Wl,-soname -Wl,$(ada_soname)'
+
ifeq (,$(findstring nocheck, $(DEB_BUILD_OPTIONS)))
CONFIGURE_OPTIONS += -DBUILD_TEST=ON
endif
--
2.19.2
Bug#914680: libplplotada3: must change the SOversion
Package: libplplotada3 Version: 5.13.0+dfsg-10 Followup-For: Bug #914680 Hello. Lintian has spotted the error leading to this situation: the shared library package named libplplotada$(soversion) is supposed to contain a file named: /usr/lib/$(multiarch)/libplplotada.so.$(soversion) but libplplotada3 currently contains /u/l/*/libplplotada.so.2.1.1. The correct fix to both problems is to rename the library. Keeping 2.1.1 makes the promise that the new library offers the same ABI to executables linked with the previous version, which is is probably false because they were built with gnat-7. You may - rename the library as libplplotada.so.3 and keep the package name - rename both the library and the package if you prefer something like 2.1.1.3 for consistency with the other languages.
Bug#913371: gnat-gps fails to build on 32-bit kernels, address space exhaustion.
Package: src:gnat-gps Followup-For: Bug #913371 Maybe we should restrict the architecture set for this binary package? I doubt that anyone is using a graphical IDE like gnat-gps on armel/mips/mipsel. On the other hand, building on all architectures improves the quality, so I would remove failing architectures instead of providing an explicit list. Any better idea?
Bug#913371: gnat-gps fails to build on 32-bit kernels, address space exhaustion.
Package: src:gnat-gps Followup-For: Bug #913371 The issue is reproducible on mipsel with MALLOC_ARENA_MAX = 1 DEB_BUILD_OPTIONS = noopt parallel=1 The package was building in experimental, the faulty version only differs by LDFLAGS += -Wl,--no-allow-shlib-undefined -Wl,--no-copy-dt-needed-entries The dependencies have not changed much either. Maybe the problem probably originates in compiler changes.
Bug#913371: gnat-gps fails to build on 32-bit kernels, address space exhaustion.
Package: src:gnat-gps Followup-For: Bug #913371 In case this helps: 18-3 managed to build on i386. https://buildd.debian.org/status/fetch.php?pkg=gnat-gps=i386=18-3=1542064776=0
Bug#913371: gnat-gps fails to build on 32-bit kernels, address space exhaustion.
Package: src:gnat-gps Followup-For: Bug #913371 Control: reopen -1 Version: 18-3 With parallelism disabled, the build should use (at least) two times less memory at a given moment. However, it fails with the same error on the same architectures. Reopening the bug. Sorry for the noise.
Bug#907988: rubber must specify the encoding when opening files
Package: rubber Followup-For: Bug #907988 Hi. I have checked that 1.5.1 fixes the build-indep target for why3.
Bug#897698: ASIS bug box in tests with gcc-8
Version 1.19r10-3 builds in experimental with an updated ASIS compatible with gnat-8. Next upload of all dependencies in unstable should fix this.
Bug#897736: dbusada: ftbfs with GCC-8
Hello. The attachment contains the same suggestions, presented as commits to the two branches for your convenience. The problem in tests was caused by ahven and has been fixed there. dbusada-commits.tar.gz Description: application/gzip
Bug#903380: libxmlada: FTBFS with unicode-data 11
Package: src:libxmlada Followup-For: Bug #903380 Hello. Updating libxmlada for a new unicode version requires a renaming of the -dev package for xmlada and all its reverse dependencies. Such migrations happen in experimental. Most packages are already renamed for unrelated reasons (gcc-8 migration and new XMLAda release). The XMLAda update should be straightforward, but will have to wait for the reupload of migrated packages to unstable, once all such changes are finished in experimental.
Bug#896570: libgnatcoll-17.0-dev must break/replace previous gnatcoll-gtk-dev package
Source: libgnatcoll Version: 17.0.2017-4~pre1 Severity: serious Tags: patch Justification: package overwrites files from another binary package This has been forgotten in last upload to experimental, is required by Ada policy, and should not reach unstable.
Bug#892148: libgtkada17.0-dev should Break/Replace previous -dev package
Source: libgtkada Followup-For: Bug #892148 I forgot the Break/Replace dependency required by Ada policy. Thanks for spotting this while the package is in experimental.
Bug#894549: debhelper: dh_usrlocal may remove a direct subdirectory of /usr/local
> use the "hashref" based autoscript?
The attached updated diff switches
from strings and sed expressions
back to arrays (again) and hashrefs (new).
> The "justdirs" is supposed to be in "opposite" order of "dirs"
> ("justdirs" are in "removal order" while "dirs" are in "creation order")
> but as far as I can see that ordering is not preserved in the patch.
They are not supposed to be in opposite order.
Actually, @justdirs contains strictly less elements than @dirs.
During the same recursive traversal,
* @dirs is appended each parent before its childs. This is the order
in which subdirectories are discovered in $tmpdir/usr/local/ during
the traversal, so it is by construction compatible with later
re-creation during install.
* @justdirs is appended all childs before their parent. The list is
constructed in the same code part actually removing the directory
from $tmpdir/usr/local/, so this order is compatible with
uninstallation
* In both lists, the childs are traversed in alphabetical order to
ensure reproducible scripts. This is unrelated with the two previous
constraints.
If this explanation seems convincing, please paste it inside comments
before the File::Find::file recursion.
> I think dh_usrlocal might be ready for some test cases to avoid future
> breakage.
The attached test.sh implements some suggested tests.
--- a/dh_usrlocal
+++ b/dh_usrlocal
@@ -77,54 +77,49 @@
if (-d "$tmp/usr/local") {
my (@dirs, @justdirs);
- find({bydepth => 1,
- no_chdir => 1,
+ find({no_chdir => 1,
+ preprocess => sub {
+ # Ensure a reproducible traversal.
+ return sort @_;
+ },
+ postprocess => sub {
+ # Uninstall, unless a direct child of /usr/local.
+ $_ = $File::Find::dir;
+ s!^\Q$tmp\E!!;
+ push @justdirs, $_ if m!/usr/local/.*/!;
+ # Remove a directory after its childs.
+ doit('rmdir', $File::Find::dir);
+ },
wanted => sub {
+ # rmdir would fail later anyways.
+ error("${File::Find::name} is not a directory")
+if not -d $File::Find::name;
+ # Install directory before its childs.
my $fn = $File::Find::name;
- if (-d $fn) {
-my $user = 'root';
-my $group = 'staff';
-my $mode = '02775';
-if (should_use_root()) {
- my $stat = stat $fn;
- $user = getpwuid $stat->uid;
- $group = getgrgid $stat->gid;
- $mode = sprintf "%04lo", ($stat->mode & 0);
- if ($stat->uid == 0 && $stat->gid == 0) {
- $group = 'staff';
- $mode = '02775';
- }
+ $fn =~ s!^\Q$tmp\E!!;
+ return if $fn eq '/usr/local';
+ if (should_use_root()) {
+my $stat = stat $File::Find::dir;
+if ($stat->uid == 0 && $stat->gid == 0) {
+ push @dirs, "$fn 02775 root staff";
+} else {
+ my $user = getpwuid $stat->uid;
+ my $group = getgrgid $stat->gid;
+ my $mode = sprintf "%04lo", ($stat->mode & 0);
+ push @dirs, "$fn $mode $user $group";
}
-
-
-
-$fn =~ s!^\Q$tmp\E!!;
-return if $fn eq '/usr/local';
-
-# @dirs is in parents-first order for dir creation...
-unshift @dirs, "$fn $mode $user $group";
-# ...whereas @justdirs is depth-first for removal.
-push @justdirs, $fn;
-doit('rmdir', $_);
- }
- else {
-warning("$fn is not a directory");
+ } else {
+push @dirs, "$fn 02775 root staff";
}
}}, "$tmp/usr/local");
- doit('rmdir', "$tmp/usr/local");
-
- my $bs = "\\"; # A single plain backslash
- my $ebs = $bs x 2; # Escape the backslash from the shell
+
# This constructs the body of a 'sed' c\ expression which
# is parsed by the shell in double-quotes
- my $dirs = join("$ebs\n", sort @dirs);
- pop @justdirs; # don't remove directories directly in /usr/local
- my $justdirs = join("$ebs\n", reverse sort @justdirs);
if (! $dh{NOSCRIPTS}) {
autoscript($package,"postinst", "postinst-usrlocal",
- "/#DIRS#/ c${ebs}\n${dirs}");
+ { 'DIRS' => join ("\n", @dirs)}) if @dirs;
autoscript($package,"prerm", "prerm-usrlocal",
- "/#JUSTDIRS#/ c${ebs}\n${justdirs}") if length $justdirs;
+ { 'JUSTDIRS' => join ("\n", @justdirs)}) if @justdirs;
}
}
}
test.sh
Description: Bourne shell script
Bug#875818: dbusada: FTBFS on arm* and alpha: Class_va_list undeclared
Package: src:dbusada
Followup-For: Bug #875818
Hello.
The attached patch fixes both issues, and removes the now unused
stdarg_h.ads.
It compiles on amd64, i386 and armel.
From: Nicolas Boulenguez <nico...@debian.org>
Subject: Remove dangling references from g++ generated code. Closes: #875818.
--- a/debian/patches/refresh-thin.diff
+++ b/debian/patches/refresh-thin.diff
@@ -20,7 +20,7 @@ Author: Nicolas Boulenguez <nico...@debian.org>
gnatmake $(GMAKE_OPTS) -Pd_bus_ada_examples -XLIBRARY_KIND=static
install: install_lib install_$(LIBRARY_KIND)
-@@ -79,9 +79,37 @@
+@@ -79,9 +79,48 @@
@rm -rf $(OBJDIR)
@rm -rf $(LIBDIR)
@rm -rf $(DOCDIR)
@@ -47,8 +47,13 @@ Author: Nicolas Boulenguez <nico...@debian.org>
+ $(CXX) --dump-ada-spec $(dbus_h) `pkg-config --cflags $(dbus_lib)`
+ # Remove unused file.
+ rm -f $@/*_dbus_syntax_h.ads
-+ # Fix an error.
-+ sed -i '/^ subtype nullptr_t is ; --/d' $@/stddef_h.ads
++ rm -f $@/stdarg_h.ads
++ # Removed a wrongly generated definition and a reference to an
++ # ungenerated type. Neither one is used in dbusada.
++ sed -i \
++ -e '/^ subtype nullptr_t is ; --/d' \
++ -e '/uu_float128/d' \
++ $@/stddef_h.ads
+ # Remove arch, path or version dependent prefixes.
+ cd $@ && for i in *dbus_dbus_*; do \
+ mv $$i `echo $$i | sed 's/.*dbus_dbus_/dbus_/'`; \
@@ -58,5 +63,11 @@ Author: Nicolas Boulenguez <nico...@debian.org>
+ sed -i '/^with Interfaces\.C; use Interfaces\.C;$$/d' \
+ $@/dbus_memory_h.ads \
+ $@/dbus_pending_call_h.ads \
-+ $@/dbus_shared_h.ads \
-+ $@/stdarg_h.ads
++ $@/dbus_shared_h.ads
++ # On arm processors, g++-7 generates a reference to starg_h.Class_va_list,
++ # but does not generate the referenced package.
++ # Remove the function using this reference, dbusada does not use it.
++ sed -i \
++ -e '/^with stdarg_h;$$/d' \
++ -e '/^ function dbus_message_\(append\|get\)_args_valist$$/{:l;N;/\n$$/d;bl}' \
++ $@/dbus_message_h.ads
Bug#875819: dbusada and anet
Hello. These two bugs are most probably caused by the regeneration from source of the thin binding during Debian builds. I will investigate them during the weekend.
Bug#872084: gnat: Prevent migration to testing until all Ada build with gnat-7.
Package: gnat Version: 7 Severity: serious Justification: migration to testing would make some Ada packages FTBFS This RC bug should prevent migration of gnat/7 to testing until all reverse dependencies are ready to build with it (because of the versioned -dev packages, most of them require more than a bin-NMU).
Bug#866355: src:adacontrol: FTBFS because of obsolete ali files in libgnat
Package: src:adacontrol Followup-For: Bug #866355 Control: reassign -1 gnat-7 Control: affects -1 ada-reference-manual adacontrol asis dh-ada-library gnat-gps gprbuild libaunit libaws libflorist libgmpada libgnatcoll libgtkada liblog4ada libncursesada libtemplates-parser libtexttools libxmlada libxmlezout music123 opentoken ahven anet dbusada pcscada polyorb spark adacgi adasockets adabrowse libalog Hello. This problem unfortunately affects every Ada package. According to the Debian policy for Ada, an Ada library should change its -dev version when the source changes, but libgnat (provided by the gnat-6 binary package, built from the gcc-6 source package) is granted an exception. The GCC source package cannot afford passing in the NEW queue once in a while for convenience of Ada users. Circumventing this specific mismatch between gnat-6 versions requires a source upload of every Ada package. This will probably wait the migration to gnat-7. Gcc-7 should now report the problem sooner, so that we at least avoid spending time to diagnose the problem. Thanks for reporting.
Bug#838483: gprbuild starts failing with "no native compiler found for language 'ada'"
Source: gprbuild Followup-For: Bug #838483 Control: affects 839209 gprbuild Upload of 2015-6 may fix some symptoms, but not those caused by #839209.
Bug#840931: libimobiledevice: diff for NMU version 1.2.0+dfsg-3.1
Control: tags 840931 + pending
Dear maintainer,
I've prepared an NMU for libimobiledevice (versioned as 1.2.0+dfsg-3.1)
and uploaded it to DELAYED/2.
Please feel free to tell me if I should delay it longer.
Regards.
diff -Nru libimobiledevice-1.2.0+dfsg/debian/changelog libimobiledevice-1.2.0+dfsg/debian/changelog
--- libimobiledevice-1.2.0+dfsg/debian/changelog 2016-06-02 18:55:15.0 +0200
+++ libimobiledevice-1.2.0+dfsg/debian/changelog 2017-01-03 01:32:36.0 +0100
@@ -1,3 +1,12 @@
+libimobiledevice (1.2.0+dfsg-3.1) unstable; urgency=medium
+
+ * Non-maintainer upload.
+ * Upstream commits replacing SSL3.0 with TLS1.0. Closes: #840931, #847977.
+Also fix related bug in GNUTLS pairing record generation, see
+https://github.com/libimobiledevice/libimobiledevice/issues/413.
+
+ -- Nicolas Boulenguez <nico...@debian.org> Tue, 03 Jan 2017 01:32:36 +0100
+
libimobiledevice (1.2.0+dfsg-3) unstable; urgency=high
* Team upload
diff -Nru libimobiledevice-1.2.0+dfsg/debian/patches/fix-ssl-version-negotiation-for-newer-versions-of-openssl.diff libimobiledevice-1.2.0+dfsg/debian/patches/fix-ssl-version-negotiation-for-newer-versions-of-openssl.diff
--- libimobiledevice-1.2.0+dfsg/debian/patches/fix-ssl-version-negotiation-for-newer-versions-of-openssl.diff 1970-01-01 01:00:00.0 +0100
+++ libimobiledevice-1.2.0+dfsg/debian/patches/fix-ssl-version-negotiation-for-newer-versions-of-openssl.diff 2017-01-03 01:32:36.0 +0100
@@ -0,0 +1,20 @@
+Description: Fix SSL version negotiation for newer versions of OpenSSL
+ Depending on the OpenSSL version (and custom distribution patches), `SSLv3_method()`
+ would return NULL on some systems and also `SSLv23_method()` fails with some older
+ iOS versions...
+Origin: upstream, https://cgit.libimobiledevice.org/libimobiledevice.git/commit/?id=13bf235cac2201747de11652cf14fe2714ca0718
+Author: David Weinstein
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=840931
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=847977
+
+--- a/src/idevice.c
b/src/idevice.c
+@@ -687,7 +687,7 @@
+ }
+ BIO_set_fd(ssl_bio, (int)(long)connection->data, BIO_NOCLOSE);
+
+- SSL_CTX *ssl_ctx = SSL_CTX_new(SSLv3_method());
++ SSL_CTX *ssl_ctx = SSL_CTX_new(TLSv1_method());
+ if (ssl_ctx == NULL) {
+ debug_info("ERROR: Could not create SSL context.");
+ BIO_free(ssl_bio);
diff -Nru libimobiledevice-1.2.0+dfsg/debian/patches/idevice-update-gnutls-code-to-support-ios-10.diff libimobiledevice-1.2.0+dfsg/debian/patches/idevice-update-gnutls-code-to-support-ios-10.diff
--- libimobiledevice-1.2.0+dfsg/debian/patches/idevice-update-gnutls-code-to-support-ios-10.diff 1970-01-01 01:00:00.0 +0100
+++ libimobiledevice-1.2.0+dfsg/debian/patches/idevice-update-gnutls-code-to-support-ios-10.diff 2017-01-03 01:32:36.0 +0100
@@ -0,0 +1,21 @@
+Description: idevice: Update GnuTLS code to support iOS 10
+ As of iOS 10 beta 4, the GnuTLS implementation idevice_connection_enable_ssl
+ needs to be updated to support TLS. Using +VERS-TLS-ALL did not work on some
+ of the devices I tested and I wasn't sure how to fix it, but +VERS-TLS1.0 is
+ working on every device I've tested: iOS 9.0.2, 10.0b4, 8.1.1, 6.0, and 3.0.
+Origin: upstream, https://cgit.libimobiledevice.org/libimobiledevice.git/commit/?id=72643b2b83990b9cf97cc84b285b30763d44a72d
+Author: Jay Freeman (saurik)
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=840931
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=847977
+
+--- a/src/idevice.c
b/src/idevice.c
+@@ -758,7 +758,7 @@
+ gnutls_certificate_client_set_retrieve_function(ssl_data_loc->certificate, internal_cert_callback);
+ #endif
+ gnutls_init(_data_loc->session, GNUTLS_CLIENT);
+- gnutls_priority_set_direct(ssl_data_loc->session, "NONE:+VERS-SSL3.0:+ANON-DH:+RSA:+AES-128-CBC:+AES-256-CBC:+SHA1:+MD5:+COMP-NULL", NULL);
++ gnutls_priority_set_direct(ssl_data_loc->session, "NONE:+VERS-TLS1.0:+ANON-DH:+RSA:+AES-128-CBC:+AES-256-CBC:+SHA1:+MD5:+COMP-NULL", NULL);
+ gnutls_credentials_set(ssl_data_loc->session, GNUTLS_CRD_CERTIFICATE, ssl_data_loc->certificate);
+ gnutls_session_set_ptr(ssl_data_loc->session, ssl_data_loc);
+
diff -Nru libimobiledevice-1.2.0+dfsg/debian/patches/series libimobiledevice-1.2.0+dfsg/debian/patches/series
--- libimobiledevice-1.2.0+dfsg/debian/patches/series 2016-06-02 18:55:15.0 +0200
+++ libimobiledevice-1.2.0+dfsg/debian/patches/series 2017-01-03 01:32:36.0 +0100
@@ -2,3 +2,6 @@
09_use_python_config.patch
local-only-sockets.patch
gnutls-api-update.patch
+fix-ssl-version-negotiation-for-newer-versions-of-openssl.diff
+idevice-update-gnutls-code-to-support-ios-10.diff
+userpref-gnutls-fix-pairing-record-generation-and-improve-error-handling.diff
diff -Nru libimobiledevice-1.2.0+dfsg/debian/patches/userpref-gnutls-fix-pairing-record-generation-and-improve-error-handl
Bug#772576: looking for a hint
Hello. In order to investigate the issue, an upstream author asks whether the path to your HOME directory contain non-ASCII characters. Thanks.
Bug#772576: taskcoach bug 772576
Hello. Please consider checking whether you can reproduce the problem - with the .deb package distributed by upstream - in a clean Debian install, if you can start a live system or a chroot. I can see no better way to start investigating this issue. However... Popcon reports 144 users of the .deb distributed by Debian. There are probably more of them either not running popcon or installing the almost identical .deb from Ubuntu. No one else has reported the same issue for 18 months. Some really run it daily, since unrelated issues have been reported meanwhile. So I intend to reduce the severity to important and allow the package to migrate to testing. important: a bug which has a major effect on the usability of a package, without rendering it completely unusable to everyone.
Bug#772576: taskcoach: Start failed
Package: taskcoach Followup-For: Bug #772576 Hello. In order to find whether the bug originates in the Debian packaging, could you please install the debian package available at http://www.taskcoach.org/download_for_linux.html then attempt to reproduce the bug? Thanks.
Bug#833134: oolite: freezes at startup
Package: oolite
Version: 1.82-2
Severity: grave
Justification: renders package unusable
Hello.
oolite/1.82-2 freezes during startup, before writing any log.
Funnily, it does not freeze anymore when started from inside gdb.
The following traces are related to oolite/1.84-1 built directly from
the VCS. It freezes in the same way, and upstream will probably be
interested in debugging only the latest version.
If I attach the executable to gdb after it is frozen, I get this backtrace:
#0 0x7fdd098ebe9c in __lll_lock_wait () from
/lib/x86_64-linux-gnu/libpthread.so.0
#1 0x7fdd098e5b92 in pthread_mutex_lock () from
/lib/x86_64-linux-gnu/libpthread.so.0
#2 0x7fdd0bd115d5 in -[NSRecursiveLock lock] (self=0x55ab20a1a8d0,
_cmd=0x7fdd0c285d90 <_OBJC_SELECTOR_TABLE+144>) at NSLock.m:308
#3 0x7fdd0be20da9 in +[NSObject(GSCleanup) leakAt:] (self=,
_cmd=, anAddress=) at NSObject+GNUstepBase.m:200
#4 0x7fdd0bd9731b in +[NSTimeZone initialize] (self=,
_cmd=) at NSTimeZone.m:1357
#5 0x7fdd0a0230d6 in __objc_install_dtable_for_class (cls=0x7fdd0c22cd40
<_OBJC_MetaClass_NSTimeZone>) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:1030
#6 0x7fdd0a024d68 in get_implementation (sel=,
class=, receiver=) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:260
#7 objc_msg_lookup (receiver=receiver@entry=0x7fdd0c22ca00
<_OBJC_Class_NSTimeZone>, op=op@entry=0x7fdd0c1a6ad0 <_OBJC_SELECTOR_TABLE+80>)
at /build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:450
#8 0x7fdd0bc733af in +[NSCalendarDate initialize] (self=,
_cmd=) at NSCalendarDate.m:372
#9 0x7fdd0a0230d6 in __objc_install_dtable_for_class (cls=0x7fdd0c1a7280
<_OBJC_MetaClass_NSCalendarDate>) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:1030
#10 0x7fdd0a024d68 in get_implementation (sel=,
class=, receiver=) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:260
#11 objc_msg_lookup (receiver=receiver@entry=0x7fdd0c1a6e60
<_OBJC_Class_NSCalendarDate>, op=op@entry=0x7fdd0c1b9090
<_OBJC_SELECTOR_TABLE+80>) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:450
#12 0x7fdd0bcadfe9 in +[NSDate initialize] (self=0x7fdd0c1b9460
<_OBJC_Class_NSDate>, _cmd=) at NSDate.m:139
#13 0x7fdd0a0230d6 in __objc_install_dtable_for_class (cls=0x7fdd0c1b9780
<_OBJC_MetaClass_NSDate>) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:1030
#14 0x7fdd0a024d68 in get_implementation (sel=,
class=, receiver=) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:260
#15 objc_msg_lookup (receiver=0x7fdd0c1b9460 <_OBJC_Class_NSDate>,
op=0x55ab1f408ff0 <_OBJC_SELECTOR_TABLE+400>) at
/build/gcc-6-SXl3Vx/gcc-6-6.1.1/src/libobjc/sendmsg.c:450
#16 0x55ab1eefc7ce in -[OOAsyncLogger startLogging] (self=0x55ab20a84ec0,
_cmd=0x55ab1f408f90 <_OBJC_SELECTOR_TABLE+304>) at
src/Core/OOLogOutputHandler.m:323
#17 0x55ab1eefc47b in -[OOAsyncLogger init] (self=0x55ab20a84ec0,
_cmd=0x55ab1f408e70 <_OBJC_SELECTOR_TABLE+16>) at
src/Core/OOLogOutputHandler.m:276
#18 0x55ab1eefbeca in OOLogOutputHandlerInit () at
src/Core/OOLogOutputHandler.m:127
#19 0x55ab1eefa8b8 in OOLoggingInit () at src/Core/OOLogging.m:602
#20 0x55ab1ef1799a in main (argc=1, argv=0x7fff61e8a998) at
src/SDL/main.m:81
A list of system calls produced by strace is attached. It ends with:
stat("/home/sandbox/.Oolite", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
stat("/home/sandbox/.Oolite/Logs", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
stat("/home/sandbox/.Oolite/Logs/Latest.log", 0x7fffe1c96320) = -1 ENOENT (No
such file or directory)
brk(0x55f4083be000) = 0x55f4083be000
mmap(NULL, 8392704, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK,
-1, 0) = 0x7fdd4eae2000
mprotect(0x7fdd4eae2000, 4096, PROT_NONE) = 0
clone(child_stack=0x7fdd4f2e1df0,
flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
parent_tidptr=0x7fdd4f2e29d0, tls=0x7fdd4f2e2700, child_tidptr=0x7fdd4f2e29d0)
= 26374
futex(0x55f40831b8d8, FUTEX_WAIT_PRIVATE, 2, NULL
strace.gz
Description: application/gzip
Bug#830333: topal: diff for NMU version 75-2.1
Control: tags 830333 + patch Control: tags 830333 + pending Dear maintainer, I've prepared an NMU for topal (versioned as 75-2.1) and uploaded it to DELAYED/10. Please feel free to tell me if I should delay it longer. Regards. diff -Nru topal-75/debian/changelog topal-75/debian/changelog --- topal-75/debian/changelog 2016-07-11 18:38:39.0 +0200 +++ topal-75/debian/changelog 2016-07-11 18:38:39.0 +0200 @@ -1,3 +1,11 @@ +topal (75-2.1) unstable; urgency=medium + + * Non maintainer upload. + * Rebuild with gnat-6 and bump ALI version. Closes: #830333. +Select gnatgcc (currently gcc-6) instead of gcc (currently gcc-5). + + -- Nicolas Boulenguez <nico...@debian.org> Mon, 11 Jul 2016 18:12:06 +0200 + topal (75-2) unstable; urgency=medium * Remove metamail from suggests list. (Closes: #665090.) diff -Nru topal-75/debian/control topal-75/debian/control --- topal-75/debian/control 2016-07-11 18:38:39.0 +0200 +++ topal-75/debian/control 2016-07-11 18:38:39.0 +0200 @@ -2,7 +2,7 @@ Section: mail Priority: optional Maintainer: Phil Brooke <p...@debian.org> -Build-Depends: debhelper (>> 6.0.0), libreadline-dev, gnat, gnat-4.9, lynx, texlive, texlive-latex-extra, texlive-fonts-extra +Build-Depends: debhelper (>> 6.0.0), libreadline-dev, gnat, gnat-6, lynx, texlive, texlive-latex-extra, texlive-fonts-extra Standards-Version: 3.9.5 Package: topal diff -Nru topal-75/debian/rules topal-75/debian/rules --- topal-75/debian/rules 2016-07-11 18:38:39.0 +0200 +++ topal-75/debian/rules 2016-07-11 18:38:39.0 +0200 @@ -32,7 +32,7 @@ dh_testdir # Add here commands to compile the package. - $(MAKE) + $(MAKE) CC=gnatgcc #/usr/bin/docbook-to-man debian/topal.sgml > topal.1 touch build-stamp diff -Nru topal-75/Makefile topal-75/Makefile --- topal-75/Makefile 2016-07-11 18:38:39.0 +0200 +++ topal-75/Makefile 2016-07-11 18:38:39.0 +0200 @@ -38,6 +38,8 @@ DEBCFLAGS=$(filter-out -Wformat,$(shell dpkg-buildflags --get CFLAGS)) +CC ?= gcc + # The default action. all: binary topal.pdf @@ -46,13 +48,13 @@ all2: package distrib ada-readline-c.o: ada-readline-c.c - gcc -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) ada-readline-c.c + $(CC) -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) ada-readline-c.c ada-echo-c.o: ada-echo-c.c - gcc -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) ada-echo-c.c + $(CC) -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) ada-echo-c.c externals-c.o: externals-c.c - gcc -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) externals-c.c + $(CC) -c -Wall $(shell dpkg-buildflags --get CFLAGS) $(TOPALDEBUG) externals-c.c mime-tool: make -C MIME-tool diff -Nru topal-75/MIME-tool/Makefile topal-75/MIME-tool/Makefile --- topal-75/MIME-tool/Makefile 2016-07-11 18:38:39.0 +0200 +++ topal-75/MIME-tool/Makefile 2016-07-11 18:38:39.0 +0200 @@ -1,9 +1,11 @@ .PHONY: all clean realclean distrib +CC ?= gcc + all: mime-tool mime-tool: mime.c - gcc $(shell dpkg-buildflags --get CFLAGS) $(shell dpkg-buildflags --get LDFLAGS) -Wall -o mime-tool mime.c + $(CC) $(shell dpkg-buildflags --get CFLAGS) $(shell dpkg-buildflags --get LDFLAGS) -Wall -o mime-tool mime.c -strip mime-tool realclean: clean
Bug#830383: adasockets: diff for NMU version 1.9-1.1
Control: tags 830383 + patch
Control: tags 830383 + pending
Dear maintainer,
I've prepared an NMU for adasockets (versioned as 1.9-1.1) and
uploaded it to DELAYED/10. Please feel free to tell me if I
should delay it longer.
Regards.
diff -Nru adasockets-1.9/configure.ac adasockets-1.9/configure.ac
--- adasockets-1.9/configure.ac 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/configure.ac 2016-07-11 18:09:01.0 +0200
@@ -7,7 +7,7 @@
#
# Package description
#
-LIBVERSIONINFO=5:0:0
+LIBVERSIONINFO=6:0:0
AC_SUBST(LIBVERSIONINFO)
AM_PROG_LIBTOOL
#
diff -Nru adasockets-1.9/debian/adasockets.gpr adasockets-1.9/debian/adasockets.gpr
--- adasockets-1.9/debian/adasockets.gpr 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/adasockets.gpr 2016-07-11 18:09:01.0 +0200
@@ -1,5 +1,5 @@
project Adasockets is
- Version := "5";
+ Version := "6";
for Library_Dir use "/usr/lib";
for Library_Name use "adasockets";
for Library_Kind use "relocatable";
diff -Nru adasockets-1.9/debian/changelog adasockets-1.9/debian/changelog
--- adasockets-1.9/debian/changelog 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/changelog 2016-07-11 18:09:01.0 +0200
@@ -1,3 +1,12 @@
+adasockets (1.9-1.1) unstable; urgency=medium
+
+ * Non maintainer upload.
+ * Rebuild with gnat-6, bump ALI and SO versions. Closes: #830383.
+Conflict with previous -dev can be implicit via gnat-X per Ada policy.
+When possible, copy versions automatically to ease next update.
+
+ -- Nicolas Boulenguez <nico...@debian.org> Mon, 11 Jul 2016 16:07:44 +0200
+
adasockets (1.9-1) unstable; urgency=medium
* New upstream release (closes: #756864).
diff -Nru adasockets-1.9/debian/control adasockets-1.9/debian/control
--- adasockets-1.9/debian/control 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/control 2016-07-11 18:09:01.0 +0200
@@ -3,14 +3,12 @@
Section: devel
Priority: extra
Standards-Version: 3.9.6
-Build-Depends: gnat, gnat-4.9, debhelper (>= 6.0.0), texinfo, texlive, dh-autoreconf
+Build-Depends: gnat, gnat-6, debhelper (>= 6.0.0), texinfo, texlive, dh-autoreconf
-Package: libadasockets5-dev
+Package: libadasockets6-dev
Architecture: any
-Depends: gnat, gnat-4.9, libadasockets5 (=${binary:Version}), ${misc:Depends}, dpkg (>= 1.15.4) | install-info
+Depends: gnat, gnat-6, libadasockets6 (=${binary:Version}), ${misc:Depends}, dpkg (>= 1.15.4) | install-info
Section: libdevel
-Conflicts: adasockets, libadasockets0-dev, libadasockets-dev, libadasockets2-dev, libadasockets3-dev, libadasockets4-dev
-Replaces: adasockets, libadasockets0-dev, libadasockets-dev, libadasockets2-dev, libadasockets3-dev, libadasockets4-dev
Description: bindings for socket services in Ada
This package provides a binding for socket services to be used with GNAT
(the GNU Ada compiler). One can use it to write clients and servers that
@@ -19,7 +17,7 @@
This package provides the Ada programmer with headers and libraries used
to build Ada programs that use sockets services.
-Package: libadasockets5
+Package: libadasockets6
Architecture: any
Depends: ${shlibs:Depends}, ${misc:Depends}
Section: libs
diff -Nru adasockets-1.9/debian/libadasockets5-dev.dirs adasockets-1.9/debian/libadasockets5-dev.dirs
--- adasockets-1.9/debian/libadasockets5-dev.dirs 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/libadasockets5-dev.dirs 1970-01-01 01:00:00.0 +0100
@@ -1,2 +0,0 @@
-usr/share/ada/adainclude/adasockets
-usr/lib/ada/adalib/adasockets
diff -Nru adasockets-1.9/debian/libadasockets5-dev.links adasockets-1.9/debian/libadasockets5-dev.links
--- adasockets-1.9/debian/libadasockets5-dev.links 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/libadasockets5-dev.links 1970-01-01 01:00:00.0 +0100
@@ -1 +0,0 @@
-usr/share/doc/libadasockets5 usr/share/doc/libadasockets5-dev
diff -Nru adasockets-1.9/debian/libadasockets5.files adasockets-1.9/debian/libadasockets5.files
--- adasockets-1.9/debian/libadasockets5.files 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/libadasockets5.files 1970-01-01 01:00:00.0 +0100
@@ -1 +0,0 @@
-usr/lib/libadasockets.so.*
diff -Nru adasockets-1.9/debian/libadasockets5.lintian-overrides adasockets-1.9/debian/libadasockets5.lintian-overrides
--- adasockets-1.9/debian/libadasockets5.lintian-overrides 2016-07-11 18:09:00.0 +0200
+++ adasockets-1.9/debian/libadasockets5.lintian-overrides 1970-01-01 01:00:00.0 +0100
@@ -1 +0,0 @@
-libadasockets5: hardening-no-relro usr/lib/libadasockets.so.5.0.0
diff -Nru adasockets-1.9/debian/libadasockets6.lintian-overrides adasockets-1.9/debian/libadasockets6.lintian-overrides
--- adasockets-1.9/debian/libadasockets6.lintian-overrides 1970-01-01 01:00:00.0 +0100
+++ adasockets-1.9/debian/libadasockets6.lintian-overrides 2016-07-11 18:09:01.0 +0200
@@
Bug#830349: adacgi: diff for NMU version 1.6-19.1
Control: tags 830349 + patch
Control: tags 830349 + pending
Dear maintainer,
I've prepared an NMU for adacgi (versioned as 1.6-19.1) and
uploaded it to DELAYED/10. Please feel free to tell me if I
should delay it longer.
Regards.
diff -Nru adacgi-1.6/debian/changelog adacgi-1.6/debian/changelog
--- adacgi-1.6/debian/changelog 2016-07-11 16:01:05.0 +0200
+++ adacgi-1.6/debian/changelog 2016-07-11 16:01:05.0 +0200
@@ -1,3 +1,10 @@
+adacgi (1.6-19.1) unstable; urgency=medium
+
+ * Non maintainer upload.
+ * Rebuild with gnat-6 and bump ALI version. Closes: #830349.
+
+ -- Nicolas Boulenguez <nico...@debian.org> Mon, 11 Jul 2016 15:39:26 +0200
+
adacgi (1.6-19) unstable; urgency=medium
* Add notice to debian/copyright re: adacgi.spec. (Closes: 756383.)
diff -Nru adacgi-1.6/debian/control adacgi-1.6/debian/control
--- adacgi-1.6/debian/control 2016-07-11 16:01:05.0 +0200
+++ adacgi-1.6/debian/control 2016-07-11 16:01:05.0 +0200
@@ -3,14 +3,12 @@
Section: devel
Priority: optional
Standards-Version: 3.9.6
-Build-Depends: debhelper (>> 6.0.0), gnat, gnat-4.9
+Build-Depends: debhelper (>> 6.0.0), gnat, gnat-6
-Package: adacgi2
+Package: adacgi3
Architecture: any
Section: web
-Depends: gnat, gnat-4.9, ${misc:Depends}
-Conflicts: adacgi, adacgi0, adacgi1
-Replaces: adacgi, adacgi0, adacgi1
+Depends: gnat, gnat-6, ${misc:Depends}
Description: Ada CGI interface
This is David A. Wheeler's Ada 95 interface to the "Common Gateway
Interface" (CGI). This makes it easier to create Ada programs that
diff -Nru adacgi-1.6/debian/rules adacgi-1.6/debian/rules
--- adacgi-1.6/debian/rules 2016-07-11 16:01:05.0 +0200
+++ adacgi-1.6/debian/rules 2016-07-11 16:01:05.0 +0200
@@ -10,7 +10,7 @@
DEB_HOST_GNU_TYPE ?= $(shell dpkg-architecture -qDEB_HOST_GNU_TYPE)
DEB_BUILD_GNU_TYPE ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_TYPE)
-ALIVERSION=2
+ALIVERSION=3
configure:
# Nothing to do.
Bug#827099: adabrowse: fails to build from source with current gnat
Package: adabrowse Severity: serious Tags: patch Justification: Fails to build from source with current GNAT Hello. The adabrowse Debian package must be rebuilt with gnat-6 and asis-2015. The attached patches fix this, and various cosmetic issues. Also, some existing bugs may probably be closed as obsolete. >From 2a389cbb44e58e573b06f84f0a601cc64ce69718 Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez <nicolas.bouleng...@free.fr> Date: Thu, 2 Jun 2016 23:58:43 +0200 Subject: [PATCH 01/20] Replace HTTP with HTTPS in VCS-* fields. --- debian/control | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/debian/control b/debian/control index 3530be7..a3c958a 100644 --- a/debian/control +++ b/debian/control @@ -5,8 +5,8 @@ Maintainer: Adrian-Ken Rueegsegger <k...@codelabs.ch> Build-Depends: debhelper (>= 9), gnat, gnat-4.9, libasis2014-dev, quilt Standards-Version: 3.9.6 Homepage: http://home.datacomm.ch/t_wolf/tw/ada95/adabrowse/ -Vcs-Browser: http://git.codelabs.ch/?p=debian/adabrowse.git -Vcs-Git: http://git.codelabs.ch/git/debian/adabrowse.git +Vcs-Browser: https://git.codelabs.ch/?p=debian/adabrowse.git +Vcs-Git: https://git.codelabs.ch/git/debian/adabrowse.git Package: adabrowse Architecture: any -- 2.8.1 >From 05aa8bc36930cc62c8d56fef46c2a3ffeda66c1c Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez <nicolas.bouleng...@free.fr> Date: Fri, 3 Jun 2016 00:05:09 +0200 Subject: [PATCH 02/20] Switch to dpkg-source 3.0 (quilt) format. --- debian/README.source | 8 debian/control | 2 +- debian/rules | 2 -- debian/source/format | 1 + 4 files changed, 2 insertions(+), 11 deletions(-) delete mode 100644 debian/README.source create mode 100644 debian/source/format diff --git a/debian/README.source b/debian/README.source deleted file mode 100644 index e35950e..000 --- a/debian/README.source +++ /dev/null @@ -1,8 +0,0 @@ -This package uses quilt to manage all modifications to the upstream -source. Changes are stored in the source package as diffs in -debian/patches and applied during the build. Please see: - -/usr/share/doc/quilt/README.source - -for more information on how to apply the patches, modify patches, or -remove a patch. diff --git a/debian/control b/debian/control index a3c958a..3870b64 100644 --- a/debian/control +++ b/debian/control @@ -2,7 +2,7 @@ Source: adabrowse Section: devel Priority: optional Maintainer: Adrian-Ken Rueegsegger <k...@codelabs.ch> -Build-Depends: debhelper (>= 9), gnat, gnat-4.9, libasis2014-dev, quilt +Build-Depends: debhelper (>= 9), gnat, gnat-4.9, libasis2014-dev Standards-Version: 3.9.6 Homepage: http://home.datacomm.ch/t_wolf/tw/ada95/adabrowse/ Vcs-Browser: https://git.codelabs.ch/?p=debian/adabrowse.git diff --git a/debian/rules b/debian/rules index b8a3398..7530d52 100755 --- a/debian/rules +++ b/debian/rules @@ -1,7 +1,5 @@ #!/usr/bin/make -f -include /usr/share/quilt/quilt.make - DPKG_EXPORT_BUILDFLAGS = 1 include /usr/share/dpkg/buildflags.mk diff --git a/debian/source/format b/debian/source/format new file mode 100644 index 000..163aaf8 --- /dev/null +++ b/debian/source/format @@ -0,0 +1 @@ +3.0 (quilt) -- 2.8.1 >From adb94b6202fd4d44ea83fcf2b4422e91ec5ea3f5 Mon Sep 17 00:00:00 2001 From: Nicolas Boulenguez <nicolas.bouleng...@free.fr> Date: Fri, 3 Jun 2016 00:06:11 +0200 Subject: [PATCH 03/20] Refreshed patches without timestamps noise. --- debian/patches/01-gnatmake.patch| 14 ++ debian/patches/02-use-gnatgcc.patch | 7 +++ 2 files changed, 9 insertions(+), 12 deletions(-) diff --git a/debian/patches/01-gnatmake.patch b/debian/patches/01-gnatmake.patch index ef86c26..522ca2f 100644 --- a/debian/patches/01-gnatmake.patch +++ b/debian/patches/01-gnatmake.patch @@ -1,10 +1,9 @@ Author: Ludovic Brenta <lbre...@debian.org> Description: Make sure never to call the "gnat" command, just "gnatmake" or "gnatgcc". -diff -Naurp adabrowse.orig/ad-driver.adb adabrowse/ad-driver.adb adabrowse.orig/ad-driver.adb 2009-02-24 21:17:01.0 +0100 -+++ adabrowse/ad-driver.adb 2009-02-24 21:17:46.0 +0100 -@@ -660,7 +660,7 @@ package body AD.Driver is +--- a/ad-driver.adb b/ad-driver.adb +@@ -660,7 +660,7 @@ -- it needs (configuration pragmas, naming schemes, etc.) -- Note: we also pass on -X options! AD.Compiler.Set_Compile_Command @@ -13,10 +12,9 @@ diff -Naurp adabrowse.orig/ad-driver.adb adabrowse/ad-driver.adb AD.Projects.Get_Project_File_Name & ASU.To_String (Pass_On_Options)); end if; -diff -Naurp adabrowse.orig/ad-projects.ads adabrowse/ad-projects.ads adabrowse.orig/ad-projects.ads 2009-02-24 21:17:01.0 +0100 -+++ adabrowse/ad-projects.ads 2009-02-24 21:17:46.0 +0100 -@@ -75,9 +75,10 @@ package AD.Projects is +--- a/ad-projects.ads b/ad-projects.ads +@@ -75,9 +7
Bug#825180: apq: fails to build from source due to gnat-6 migration
Source: apq
Version: 3.2.0-2
Severity: serious
Tags: patch
Justification: FTBFS
Hello.
Because of the Debian gnat-6 migration, the APQ package currently
fails to build from source.
The attached patch also fixes various less important issues.
From: Nicolas Boulenguez <nico...@debian.org>
Date: Tue, 24 May 2016 13:18:16 +0200
Subject: suggestion for APQ Debian package
* Rebuild with gnat-6. Rename lib and -dev packages per Ada policy.
Update ALI version, SO version and -dev dependencies accordingly.
* Switch to gprbuild because gnatmake is dropping support for GNAT projects.
* library-options.patch: most LDFLAGS need to be listed before -l options.
* Switch from manual -dbg package to automatic -dbgsym package.
* Enable all hardening flags.
* Update Homepage, Standards-Version, Testsuite.
Note: even if diff does not encode this properly, the
debian/libapq*.install files are renamed, not modified.
diff --git a/debian/control b/debian/control
index 8184299..2d687f0 100644
--- a/debian/control
+++ b/debian/control
@@ -4,16 +4,19 @@ Priority: optional
Maintainer: Adrian-Ken Rueegsegger <k...@codelabs.ch>
Uploaders: Reto Buerki <r...@codelabs.ch>,
Nicolas Boulenguez <nico...@debian.org>
-Build-Depends: debhelper (>= 9), gnat, gnat-4.9
+Build-Depends: debhelper (>= 9), gnat, gnat-6,
+ dpkg-dev (>= 1.16.1),
+# providing /usr/share/dpkg/buildflags.mk
+ gprbuild (>= 2015-2)
+# handling gnat-6
Build-Depends-Indep: texlive-latex-recommended, texlive-fonts-recommended,
latex-xcolor, texlive-latex-extra
-Standards-Version: 3.9.5
-Homepage: https://bitbucket.org/kowframework/
+Standards-Version: 3.9.8
+Homepage: https://github.com/ada-apq/apq/
Vcs-Browser: http://git.codelabs.ch/?p=debian/apq.git
Vcs-Git: http://git.codelabs.ch/git/debian/apq.git
-XS-Testsuite: autopkgtest
-Package: libapq3.2.0.1
+Package: libapq3.2.0.2
Architecture: any
Depends: ${shlibs:Depends}, ${misc:Depends}
Description: Pluggable Ada 95 Binding to various database systems (library)
@@ -21,12 +24,10 @@ Description: Pluggable Ada 95 Binding to various database systems (library)
PostgreSQL or MySQL. This is the core package providing the plugin
loading support.
-Package: libapq4-dev
+Package: libapq5-dev
Section: libdevel
Architecture: any
-Depends: ${misc:Depends}, libapq3.2.0.1 (= ${binary:Version}), gnat, gnat-4.9
-Breaks: libapq3.2.0-dev
-Replaces: libapq3.2.0-dev
+Depends: ${misc:Depends}, libapq3.2.0.2 (= ${binary:Version}), gnat, gnat-6
Description: Pluggable Ada 95 Binding to various database systems (development)
APQ is a pluggable Ada 95 Binding to various database systems like
PostgreSQL or MySQL. This is the core package providing the plugin
@@ -34,20 +35,6 @@ Description: Pluggable Ada 95 Binding to various database systems (development)
.
This package contains the development files.
-Package: libapq-dbg
-Section: debug
-Priority: extra
-Architecture: any
-Depends: ${misc:Depends}, libapq3.2.0.1 (= ${binary:Version})
-Recommends: libapq4-dev (= ${binary:Version})
-Suggests: gnat
-Description: Pluggable Ada 95 Binding to various database systems (debug)
- APQ is a pluggable Ada 95 Binding to various database systems like
- PostgreSQL or MySQL. This is the core package providing the plugin
- loading support.
- .
- This package contains the debugging symbols.
-
Package: libapq-doc
Section: doc
Architecture: all
diff --git a/debian/copyright b/debian/copyright
index 07d0889..9134cf2 100644
--- a/debian/copyright
+++ b/debian/copyright
@@ -1,7 +1,7 @@
Format: http://www.debian.org/doc/packaging-manuals/copyright-format/1.0/
Upstream-Name: APQ
Upstream-Contact: Marcelo Coraça de Freitas <marc...@kow.com.br>
-Source: http://bitbucket.org/kowframework/apq/src
+Source: https://github.com/ada-apq/apq/
Files: *
Copyright:
@@ -35,7 +35,7 @@ License: GPL-2+ with Ada exception
Files: debian/*
Copyright:
(C) 2009-2014 Adrian-Ken Rueegsegger <k...@codelabs.ch>
- (C) 2014 Nicolas Boulenguez <nico...@debian.org>
+ (C) 2014-2016 Nicolas Boulenguez <nico...@debian.org>
License: GPL-2+
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
diff --git a/debian/libapq3.2.0.1.install b/debian/libapq3.2.0.1.install
deleted file mode 100644
index 49cc562..000
--- a/debian/libapq3.2.0.1.install
+++ /dev/null
@@ -1 +0,0 @@
-/usr/lib/*.so.*
diff --git a/debian/libapq3.2.0.2.install b/debian/libapq3.2.0.2.install
new file mode 100644
index 000..49cc562
--- /dev/null
+++ b/debian/libapq3.2.0.2.install
@@ -0,0 +1 @@
+/usr/lib/*.so.*
diff --git a/debian/libapq4-dev.install b/debian/libapq4-dev.install
deleted file mode 100644
index c733a12..000
--- a/debian/libapq4-dev.install
+++ /dev/null
@@ -1,4 +0,0 @@
-/usr/lib/*.a
-/usr/lib/*.so
-/usr/lib/ada
-/usr/share/ada
diff --git a/debian/libapq5-dev.install b/debian/libapq5-dev.install
new file mode 100644
index 000
Bug#823369: adacontrol: FTBFS: new ARM restrictions for representation clauses
Package: src:adacontrol Followup-For: Bug #823369 Hello. The log attached by Matthias contains the same compiler message, though the problem is detected far later. New restrictions make legal Ada95 code illegal in Ada2005 [1]. All Debian architectures ensure "sizeof(long long) == 8" [2] probably implying "the size of the largest machine scalar is at least 64". Adding the Ada_2005 and lowering the length of the RecTab string field to 7 chararcters instead of 10 fixes the issue on partch.debian.org. [1] http://www.ada-auth.org/cgi-bin/cvsweb.cgi/ais/ai-00133.txt. [2] https://wiki.debian.org/ArchitectureSpecificsMemo,
Bug#772576: taskcoach: Start failed
Package: taskcoach Followup-For: Bug #772576 Hello. The issue may be grave, but it should not block taskcoach forever if noone is in a position to investigate it. Please consider: - attaching the contents of "~/.config/Task Coach/TaskCoach.ini" In case you have activated the synchronization with an iPhone, be sure to remove your password from these contents. - running "LANG=C python -mtrace —trace /usr/bin/taskcoach > /tmp/log.txt 2>&1" inside a terminal, then attach the contents of "/tmp/log.txt". Thanks.
Bug#777934: libalog: ftbfs with GCC-5
Package: src:libalog Followup-For: Bug #777934 Hello. You may be interested in the last commit of branch org.debian.libgnatcoll in the Ada France monotone repository, closing the exact same issue.
Bug#777934: libalog: ftbfs with GCC-5
Package: src:libalog Followup-For: Bug #777934 Hello. Gprbuild, by default, only considers compilers - which are the Debian default (ex: /usr/bin/gcc from package gcc) - for which the version starts with 4.9. The intent is that users of the stable distribution with default settings - use the Debian default compilers - never mix objects with incompatible ABIs. Consequently, upgrading gcc (Debian package providing the /usr/bin symlink) to version 5 breaks libalog (#777934). The relevant line of the log ends with warning: no compiler specified for language C, ignoring all its sources Gprbuild detects no C compiler (gcc-4.9 is installed as a Build-Dependency, but /usr/bin/gcc --version does not answer 4.9), ignores all C sources, and later complains at link stage about missing symbols. Anyway, due to Debian Ada policy, libalog cannot be built with gnat-5 before libahven. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#777946: libgnatcoll: ftbfs with GCC-5
Package: src:libgnatcoll Followup-For: Bug #777946 Hello. Gprbuild, by default, only considers compilers - which are the Debian default (ex: /usr/bin/gcc from package gcc) - for which the version starts with 4.9. The intent is that users of the stable distribution with default settings - use the Debian default compilers - never mix objects with incompatible ABIs. Consequently, upgrading gcc (Debian package providing the /usr/bin symlink) to version 5 breaks libgnatcoll (#777946). The relevant line of the log ends with warning: no compiler specified for language C, ignoring all its sources Gprbuild detects no C compiler (gcc-4.9 is installed as a Build-Dependency, but /usr/bin/gcc --version does not answer 4.9), ignores all C sources, and later complains at link stage about missing symbols. Anyway, due to Debian Ada policy, libgnatcoll cannot be built with gnat-5 before libgtkada. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#778039: Patch for GCC 5 build issue
Package: src:oolite Followup-For: Bug #778039 Here's a fix for the GCC 5 build issue. I added -Wno-error=format-security I don't think that's the best way forward. Since this is security-relevant I think we should rather fix the problem than ignoring the errors. Fully agreed. However, thanks reporting The package successfully builds and links using GCC 5 with this change. because it shows that gcc-5 will not expose any other problem. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#777946: libgnatcoll: ftbfs with GCC-5
Package: src:libgnatcoll Followup-For: Bug #777946 When compiling both Ada and C sources, gprbuild default knowledge base requires gnat=gnat-4.9 and gcc=gcc-4.9, ensuring that the objects can safely be linked together with the default compilers. When gcc=gcc-5, gprbuild fails to see gcc-4.9 and ignores C sources. This results in undefined symbols at the link stage. The problem may probably be circumvented by specifying an hand-crafted gprbuild knowledge base during libgnatcoll build. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#788974: taskcoach and lockfile Debian bug 788974
Package: taskcoach Followup-For: Bug #788974 Hello. TaskCoach embeds a copy of python-lockfile-0.8 sources, patches them at build time, then installs the patched sources. Debian packaging replaces the copy with the python-lockfile Debian package, then lets the upstream's Makefile patch and install it. Python-lockfile 0.10 has broken all this, and adapting the patch is far from trivial. I have split, documented and forwarded the patch to lockfiles's upstream, see https://launchpad.net/pylockfile. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#768547: gnat-gps-doc: wrong license for user guide (GPL instead of GFDL)
Source: gnat-gps Followup-For: Bug #768547 Control: tags -1 pending I have modified the com.adacore.gps.debian and org.debian.gnat-gps branches of the VCS so that the new license is unchanged and mentioned in debian/copyright. To minimize changes during the freeze, I have not removed the gnatlib/ subdirectory. The package is waiting for build reverse dependencies to be checked for installability on all architectures (dh-ada-library may have marked lib*-dev as depending on gnat-4.9.1, especially libgtkada-dev recently), but I think the issue is solved. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#768805: dh-ada-library: Must accept output of recent gnatmake --version
Package: dh-ada-library Version: 6.2 Severity: serious Justification: creates wrong dependencies across binary packages Dh-ada-library relies on the output of gnatmake --version. The format of this message has changed with gnat-4.9 = 4.9.1-4. As a consequence, wrong debhelper substitution variables are generated, containing gnat-4.9.1 instead of gnat-4.9. Reverse build-dependencies will build binary packages with non-existing dependencies. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#730893: oolite: FTBFS: [Java framework] Error in function createSettingsDocument (elements.cxx).
Package: src:oolite Followup-For: Bug #730893 Hello. Would someone please attempt to build the attached source package with sbuild? It only demonstrates a minimal conversion from odt to PDF format with libreoffice, and works on my pbuilder. If it fails, I suggest to reassign the bug to libreoffice, maybe marking that oolite is affected. Else, I intend to solve the issue for Jessie by avoiding the conversion at all, and distributing the unchanged .odt files instead of the PDF. I would let the bug open for later investigation, but we may then all agree to downgrade its severity to minor. Format: 3.0 (native) Source: reproducer Binary: reproducer Architecture: all Version: 1 Maintainer: Nicolas Boulenguez nico...@debian.org Standards-Version: 3.9.6 Build-Depends: debhelper (= 9), libreoffice-writer Package-List: reproducer deb games optional arch=all Checksums-Sha1: 79195703db6095382e1e29e11f88fa0951ae955a 7800 reproducer_1.tar.xz Checksums-Sha256: 2352b207645d94b51c52c43e971b2285ed47295158a3c4817ef37088d4905fb0 7800 reproducer_1.tar.xz Files: a08d1ab1fde84d3ff33429259b1ec262 7800 reproducer_1.tar.xz reproducer_1.tar.xz Description: application/xz
Bug#730893: builds okay in cowbuilder on amd64
On Sun, Nov 09, 2014 at 07:57:01PM +0100, Lucas Nussbaum wrote: Why don't you just fix it by creating a temporary directory, and setting $HOME to it? For quite a good reason: because I did not think to this by myself. In my previous answer, please replace avoiding the conversion at all, and distributing the unchanged .odt files instead of the PDF. with creating a temporary directory, and setting $HOME to it. Thanks. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#708942: [gprbuild] GFDL non free file with front cover
Source: gprbuild Followup-For: Bug #708942 Control: reopen -1 ! A recent clarification [1] by upstream authors makes it explicit that the documentation *is* licensed under the GFDL with invariant sections, and must be removed from Debian. [1] https://lists.debian.org/debian-ada/2014/10/msg00023.html -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#768547: gnat-gps-doc: wrong license for user guide (GPL instead of GFDL)
Package: gnat-gps-doc Severity: serious Justification: license issue The file docs/users_guide/license.rst in the pristine upstream tarball contains the GFDL. It is included by docs/users_guide/index.rst, which mentions no invariant parts or cover texts, so the documentation complies to the DFSG. However.. Upstream authors used to claim that all the whole archive was licensed under the GPL, though some obsolete boilerplates were telling a different story. The Debian .orig archive used to be repackaged to clarify this issue. A recent statement [1] by upstream authors makes it explicit that the documentation *is* licensed under the GFDL. [1] https://lists.debian.org/debian-ada/2014/10/msg00023.html It seems that gnat-gps must be repackaged again with the correct licensing. Maybe this is the opportunity to also remove the embedded copy of libgnatcoll lying in the gnatlib/ subdirectory (see debian/TODO). -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#768597: libgtkada: GFDL non free file with front cover
Source: libgtkada Severity: serious Justification: license issue A recent clarification [1] by upstream authors makes it explicit that files under docs/gtkada_ug/ *are* licensed under the GFDL with invariant sections, and must be removed from Debian. [1] https://lists.debian.org/debian-ada/2014/10/msg00023.html -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#768214: #768214
Package: gprbuild-doc Followup-For: Bug #768214 Control: tag -1 pending This issue is fixed in the version control system, but I intend to also work on the license problem described at https://lists.debian.org/debian-ada/2014/11/msg0.html before uploading in order to limit the number of unblock requests. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#762413: taskcoach: Please relax dependency on python-wxgtk3.0
Package: src:taskcoach Followup-For: Bug #762413 After my explanations and the upload of taskcoach_1.4.1-2, which depends on python-wxgtk3.0 3.0.1.1+dfsg.0~ and = 3.0.1.1+dfsg, I consider that a serious severity is not justified by blocks an on-going transition anymore. Do you agree to close this bug, or downgrade it to wishlist as a reminder that a more satisfying solution for pubsub has yet to be found? -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#763284: libgnatcoll: GPL code linking with OpenSSL via PostGreSQL
Source: libgnatcoll Severity: serious Justification: Policy 2.1 GPL licenses are incompatible with OpenSSL. [α] summarizes the problem, with a link to Debian position and possible solutions [β]. The libgnatcoll source package must stop building the postgresql backend, because it links with OpenSSL, via PostGreSQL. I doubt that AdaCore will be interested in granting a specific exception to the GPL, and no dependency requires the PostGreSQL backend as far as I known, so the best solution is probably to stop building the libgnatcoll-postgresql[SO | ALI-dev | -dbg] binary packages. [α] https://people.gnome.org/~markmc/openssl-and-the-gpl.html [β] https://lists.debian.org/debian-legal/2002/10/msg00113.html -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#763286: apq-postgresql: GPL code links with OpenSSL via postgresql
Source: apq-postgresql Severity: serious Justification: Policy 2.1 GPL licenses are incompatible with OpenSSL. [α] summarizes the problem, with a link to Debian position and possible solutions [β]. The apq-postgresql package depends on postgresql, so must link with OpenSSL. Debian cannot distribute such binaries. According to Adrian-Ken Rueegsegger, the probability of a license change is very low, so apq and apq-postgresql will probably be removed, while the Alog reverse dependency drops PostGreSQL support. [α] https://people.gnome.org/~markmc/openssl-and-the-gpl.html [β] https://lists.debian.org/debian-legal/2002/10/msg00113.html -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#762413: taskcoach: Please relax dependency on python-wxgtk3.0
On Mon, Sep 22, 2014 at 05:20:11PM +1200, Olly Betts wrote: | Depends: python-wxgtk3.0 ( 3.0.0.0+dfsg.0~), python-wxgtk3.0 (= 3.0.0.0+dfsg), [...] The = restriction is pointless - every uploaded version satisfies it as the first upload was 3.0.0.0+dfsg-1: https://packages.qa.debian.org/w/wxpython3.0.html By adding needless versioning of dependencies like this, you're mainly just making life more difficult for the dependency resolver by giving it more constraints to process. If there's a reason for the upper bound on the version of python-wxgtk3.0, you should really have discussed that before adding it, rather than letting it be a nasty surprise now, when it blocks testing migration of a newer wxpython3.0 upload. The versioned dependency is generated by dh_linktree. Upstream embeds a copy of the pubsub source tree because it needs to patch one line in one source. I use the sources from wxgtk3.0 instead, apply the patch then call dh_linktree deduplicate to replace unpatched files with symlinks. Whether I use dh_linktree or not, I want to avoid duplicating the pubsub tree so I need to rebuild taskcoach (and reapply the patch) each time wxgtk3.0 changes its upstream version, in case the pubsub code has been modified. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759407: gnat-4.9: gnat1 not found on kfreebsd-i386
Package: src:gnat-4.9 Followup-For: Bug #759407 Hello. The success on 2014/08/11 was using gcc-4.9_4.9.1-4. The failure on 2014/08/24 was using gcc-4.9_4.9.1-7. The failure is reproducible on fischer.debian.org, with gcc-4.9_4.9.1-13 and the same gnat-4.9_4.9.1-1: # find /usr/lib -name gnat1 /usr/lib/gcc/i486-kfreebsd-gnu/4.9/gnat1 # touch p.adb gnatmake p.adb gcc-4.9: error trying to exec 'gnat1': execvp: No such file or directory -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759407: gnat-4.9: gnat1 not found on kfreebsd-i386
Package: src:gnat-4.9 Followup-For: Bug #759407 Svante's suggestion seems to work for the binutils package: it installs /usr/lib/i486* symlinks to make the /usr/lib/i586* tools visible to callers. A bootstrap issue will remain: building a new gnat-4.9 package installing these symlinks requires a working compiler. In case this helps finding a better solution: -- Where does gcc search for gnat1? gcc-4.9-source/gcc_tarball/gcc/gcc.c searches gnat1 in static const char *spec_machine = DEFAULT_TARGET_MACHINE; gcc-4.9-source/gcc_tarball/gcc/Makefile.in sets DEFAULT_TARGET_MACHINE from @target_noncanonical@. gcc-4.9-source/gcc_tarball/gcc/configure.ac calls ACX_NONCANONICAL_TARGET in config/acx.m4, which set target_noncanonical from $target_alias or $host_noncanonical if blank. gcc-4.9-source/gcc_tarball/debian/rules2 sets in CONFARGS --build=$(DEB_BUILD_GNU_TYPE) --host=$(DEB_HOST_GNU_TYPE) --target=$(TARGET_ALIAS) -- Where is gnat1 actually installed? gnat-4.9/debian/rules.d/binary-ada.mk installs gnat1 under ../$(gcc_lexec_dir)/... gnat-4.9/debian/rules2 sets gcc_lexec_dir to ../$(TARGET_ALIAS)/.. -- Common part. The commented stanzas seem to show that a similar problem has been encountered in the past. gnat-4.9/debian/rules.defs: ifeq ($(DEB_CROSS),yes) DEB_TARGET_ALIAS ?= $(DEB_TARGET_GNU_TYPE) TARGET_ALIAS := $(DEB_TARGET_ALIAS) else TARGET_ALIAS := $(DEB_TARGET_GNU_TYPE) ifeq ($(TARGET_ALIAS),i386-gnu) TARGET_ALIAS := i586-gnu endif #ifeq ($(TARGET_ALIAS),i486-linux-gnu) # TARGET_ALIAS := i686-linux-gnu #endif TARGET_ALIAS := $(subst i386,i486,$(TARGET_ALIAS)) # configure as linux-gnu, not linux #ifeq ($(findstring linux,$(TARGET_ALIAS))/$(findstring linux-gnu,$(TARGET_ALIAS)),linux/) # TARGET_ALIAS := $(TARGET_ALIAS)-gnu #endif # configure as linux, not linux-gnu #TARGET_ALIAS := $(subst linux-gnu,linux,$(TARGET_ALIAS)) -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#750261: ada-reference-manual: FTBFS: Texinfo errors
Package: src:ada-reference-manual Followup-For: Bug #750261 Control: tags -1 + unreproducible Control: severity -1 important As archive-wide rebuilds do not seem to be able to reproduce the failure either, I consider that the severity is only important. http://aws-logs.debian.net/ftbfs-logs/2014/06/28 http://aws-logs.debian.net/ftbfs-logs/2014/07/18 -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
