Source: gcc-6
Package: gnat-6
Version: 6.3.0-4
Followup-For: Bug #839209
Hello.
gcc Depends: gcc-6, and they provide the following /usr/bin symlinks.
gcc: gcc-> gcc-6
gcc: ARCH-gcc-> gcc-6
gcc-6 : gcc-6 -> ARCH-gcc-6
For gnat-6 (the gnat package has a d
Hello.
As far as I understand the situation, disabling the whole patch at
https://anonscm.debian.org/viewvc/gcccvs/branches/sid/gcc-7/debian/patches/ada-default-project-path.diff?view=markup
was the right choice for gnat-7.
It may probably be dropped, as gprbuild is already patched to search
proj
Package: gnat-7
Severity: wishlist
Hello.
Please ignore file mtimes greater than SOURCE_DATE_EPOCH when
generating timestamps into ALI files. This would help reproducible
builds of many Ada packages patching or generating Ada sources.
A first patch/test attempt is attached, but I am unable to r
(`|sed instead of |sed`, and only spaces instead of blank and
spaces, see 814978).
.
These changes have not been tested.
Author: Nicolas Boulenguez
--- a/debian/copyright.in
+++ b/debian/copyright.in
@@ -48,7 +48,6 @@ gcc-@BV@-source The sources with patches
Ada:
libgnatvsn-dev
Package: gnat-7
Followup-For: Bug #856042
Control: tags 856042 + patch
This new source-date-epoch.diff applies to 7-20170226-1 and builds
with DEB_BUILD_OPTIONS="lang=ada nocheck noopt nostrap nolang=...".
The test is more extensive, but I have not been able to run it on the
build result (for unre
egenerated from the updated copyright.in.
.
Remove libgnatvsn7.overrides, currently unused and replaced with a
more simple solution once libgnatvsn is restored.
Author: Nicolas Boulenguez
--- a/debian/copyright.in
+++ b/debian/copyright.in
@@ -48,7 +48,6 @@ gcc-@BV@-source The so
Package: gnat-7
Followup-For: Bug #856042
Thank you for applying the patch.
The test script passes (with "V=v;export V;c" instead of "V=v c").
Hello.
gcc-7 build process may recreate src/Makefile.in from src/Makefile.def
and src/Makefile.tpl with something like
# cd src && autogen Makefile.def
right after patch application.
This would drop two thousand lines of noise in
debian/patches/bootstrap-no-unneeded-libs.diff
debian/patches/g
> > # cd src && autogen Makefile.def
> > This would drop two thousand lines of noise in
> > debian/patches/bootstrap-no-unneeded-libs.diff
> > debian/patches/gdc-libphobos-build].diff
> > plus a bit more in the Ada patches.
Actually, around 1600 with future libgnatvsn patches.
> > - Gcc-7 alre
Package: src:gcc-7
Followup-For: Bug #857606
With the attachements...
>From 4562c92c5dce6831d75c3b9753cbc19f1351f1a5 Mon Sep 17 00:00:00 2001
From: Nicolas Boulenguez
Date: Mon, 13 Mar 2017 11:37:23 +0100
Subject: [PATCH 15/18] Replace manual -dbg packages with automatic -dbgsym
packa
Package: src:gcc-7
Followup-For: Bug #857606
Here is another update for the Ada patches (attachement 17).
It will simplify a lot the maintenance.
It should not modify the build process, but I have not tested.
While working on this...
16 and 18 are trivial.
15 is only a suggestion.
Matthias Klose:
> Package: src:gcc-7
> https://buildd.debian.org/status/fetch.php?pkg=gcc-7&arch=arm64&ver=7-20170314-1&stamp=1489581110&raw=0
> ali.o: file not recognized: File truncated
> seen with local builds as well for other files in the gnatmake, gnatlink
> targets.
The construction by Mak
Package: src:gcc-7
Followup-For: Bug #857831
Control: tags -1 + patch
Hello.
Please consider the attached suggestions for the GCC/Ada packaging.
"Remove ada-gnattools-noparallel.diff, should be fixed." has been
applied for several successul builds, so this bug is probably fixed.
"Reduce diff wi
also #includes , probably for historical
+ reasons since no termio structure is ever used.
+ .
+ Drop the #include instead of maintaining a list of architectures.
+Author: Nicolas Boulenguez
+Bug-Debian: https://bugs.debian.org/845159
+
+--- a/src/gcc/ada/terminals.c
b/src/gcc/ada/terminals.c
uot; \
- "ADA_INCLUDES=-I- -I../rts $(ADA_INCLUDES_FOR_SUBDIR)"\
--- a/debian/patches/ada-libgnatvsn.diff
+++ b/debian/patches/ada-libgnatvsn.diff
@@ -46,7 +46,7 @@
+} | tee -a config.log
--- /dev/null
+++ b/src/libgnatvsn/Makefile.in
-@@ -0,0 +1,156 @@
+@@ -0,0 +1,150 @@
+# Makefile
Package: src:gcc-7
Followup-For: Bug #857831
Control: tags -1 + patch
The attached patch fixes the new race condition introduced by previous
attempt.
--- a/debian/patches/ada-gnattools-cross.diff
+++ b/debian/patches/ada-gnattools-cross.diff
@@ -247,23 +247,28 @@ Index: b/src/gnattools/Makefile.in
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
>From this bug log and title, it is not clear for me whether
7-20170314-1 fixes the issue: ada-gcc-name.diff modifies gnatmake so
that it calls gcc-BV.
Are the other problems with tests fixed by these changes?
Should this bug be closed beca
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
The workaround for 814978 is a symbolic link
/usr/bin/gcc-7-7 -> gcc-7
Is it still required?
Package: src:gcc-7
Followup-For: Bug #861734
Please build the native gnat on armel by reverting
svn://anonscm.debian.org/gcccvs/branches/sid/gcc-7@9434
6ca36cf4-e1d1-0310-8c6f-e303bb2178ca
Since then, the link step for the tools has changed a lot, and a
recent build log would be more helpful. Mor
.
+Author: Nicolas Boulenguez
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=861735
+Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=861737
+
+--- a/src/gcc/ada/gcc-interface/Makefile.in
b/src/gcc/ada/gcc-interface/Makefile.in
+@@ -1424,7 +1424,7 @@
+ s-tpopsp.adb
Package: gnat-5
Followup-For: Bug #814977
Control: found -1 5.4.1-11
Related to #814978 and #856274.
Here is an easy way to reproduce.
# echo 'package P is end P;' > p.a
# /usr/bin/gnatchop p.a
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
Small reproducer:
# 'package P is end P;' > p.a
# /usr/bin/gnatchop p.a
On gnat-5/5.4.1-11, the work-around fixes the issue.
Without it, gcc-5-5 is searched but not found.
On gnat-6/6.3.0-18, the work-around has no effect.
ARCH-gcc-6-6 is s
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
On gnat-7/7.1.0-7, the work-around has no effect.
ARCH-gcc-7-7 is searched but not found.
Package: gnat-7
Severity: minor
Hello.
The fix for #856042 lets gnat write a fake time stamp into the .ali
file when an Ada source is more recent than SOURCE_DATE_EPOCH.
If another source depending on this one is later compiled with
gnatmake, gnatmake will (sensibly) conclude that the dependency
Package: gnat-5
Followup-For: Bug #806760
Control: retitle -1 [fixed in gnat-7] gnat-5 tools linked statically when built
with DEB_BUILD_OPTIONS=parallel=1
Hello.
The build process for gnat tools has changed a lot in gnat-7.
The logs show that they now link dynamically (except the gcc driver,
but
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
> On gnat-7/7.1.0-7, the work-around has no effect.
> ARCH-gcc-7-7 is searched but not found.
> what do you mean? that this doesn't work for the cross compilers?
No, simply that on a fresh (x86_64-linux-gnu) chroot:
# apt-get install gnat-
Package: src:gcc-7
Followup-For: Bug #861734
Control: retitle -1 libgnat fails to link without -latomic on armel
A rebuild with --as-needed confirms that libgnat requires libatomic,
not the tools.
As far as I understand, the problem may exist upstream, but be hidden
at build time without -z defs,
Package: gnat-7
Followup-For: Bug #866029
Control: tags -1 + pending
Now with the attachment.
A patch is in the VCS.
#!/bin/sh
set -C -e -f -u
# On a normal installation.
gnatmake=gnatmake
# Alongside a fresh gcc build tree.
# gcctree=..
# target_noncanonical=arm-linux-gnueabi
# exedir=$gcctree/
Package: gnat-7
Followup-For: Bug #864969
Control: retitle -1 Ada.Locales ignores setting of LANG
Control: tags -1 + upstream
Control: forwarded -1 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81243
Ada.Locales.Country_Code is a thin wrapper to locales.c.
void c_get_country_code (char4 p) {
cha
Package: gnat-5,gnat-6,gnat-7
Followup-For: Bug #856274
Control: retitle -1 gnat-BV: gnatchop fails to execute (ARCH-)?gcc-BV-BV
Control: affects -1 libgtkada
Work-around: Build-Depend (or rebuild with): gnat-5 (>= 5.4.1-11)
or Build-Depend (or rebuild with): gnat-6 (>= 6.3.0-18)
Source: gcc-7-cross
Severity: wishlist
Hello.
As discussed on the debian-ada mailing list, building a cross
libgnatvsn is a waste of efforts.
Version 7.1.0-11 of gcc-7 will only produce a native libgnatvsn.
The attached diff starts the work of adapting gcc-7-cross.
--- a/debian/control
+++ b/debia
uires -lrt.
It will be ignored on other platforms thanks to --as-needed.
Author: Ludovic Brenta
Author: Nicolas Boulenguez
--- a/src/gcc/ada/libgnarl/s-osinte__kfreebsd-gnu.ads
+++ b/src/gcc/ada/libgnarl/s-osinte__kfreebsd-gnu.ads
@@ -45,6 +45,7 @@ package System.OS_Interface is
pragma Pre
Hello.
Is the existing build system (in gcc-9 patches) sufficient to close
this bug, or are full configure.ac and Makefile.in better?
Hello.
It seems that debian/patches/ada-gcc-name.diff in gnat-8/8.3.0-6
applies your suggestion.
Do you confirm that this bug can be closed?
Control: reassign 717014 gnat-9 9.1.0-8
Control: reassign 760211 gnat-9 9.1.0-8
Checked on armel.
Gnatmake does not support projects anymore, a new reproducer follows.
BV=9
# Reproducer for 717014
echo 'with Ada.Numerics.Generic_Elementary_Functions;' > a.ads
echo 'package A is new Ada.Numerics.G
> 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
@@ -
> You may also contact Nicolas, who wanted to coordinate building of
> targets which are not Debian ports.
I have only suggested to share experience and possibly code. The
advices I have received from you and others may be found at
https://wiki.debian.org/PackagingLessCommonBinutilsTargets
binut
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 failu
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
Package: src:gcc-9
Followup-For: Bug #954681
Control: reassign -1 gnat-10
Control: retitle -1 CONSTRAINT_ERROR : uintp.adb:2002 overflow check failed
Since version 10-20200410-1, the SOURCE_DATE_EPOCH patch should not
trigger this bug anymore. However, this bug is most probably not
fixed, and a r
Package: gnat-9
Followup-For: Bug #748297
Hello.
The patch (now named ada-libgnat_util.diff) has improved its
integration in the GCC packaging.
Its header explains why upstream is not interested in applying it
(summary: upstream links GNAT tools with libgnat_util statically
during bootstrap, and
Control: reassign 717014 gnat-10 10.2.0-9
Control: reassign 760211 gnat-10 10.2.0-9
Reproduced on armel.
Hello.
> check_ali_update.sh is triggered with gcc-12 12.3.0-2.
Probably, the only fix requires to rename of Ada -dev packages:
~20 trivial NMUs + NEW + stable-proposed-updates
> There were no changes to gcc/ada and libada.
The change may be caused by generated code interfacing with C.
> Pleas
> check_ali_update.sh is triggered with gcc-12 12.3.0-2.
Hello.
This is a false alert.
Commit 8fd9efdd6cc7fd537aa3ac1d4c57178ca0d3e6b7 in the GCC packaging
should prevent it.
Here are some details, in case further investigation is required.
Between gnat-12_12.2.0-14_amd64.deb and gnat-12_12.3.0
Package: gnat-12
Followup-For: Bug #643663
Control: reassign 643663 gnat-13 13.2.0-5
Control: retitle 643663 gnat: invalid access to atomic variable
Hello.
With gnat-13, the following sources compile without error.
On the other hand, the invalid access to the atomic variable may not
be fixed.
Do
Hello.
A bug affects the Ada tests during the build of the gcc-12 source
package.
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1066052
Such error messages happen on each (uncoordinated) change in the sources
of an Ada library. Some sources for the standard library built by GCC
are gen
Package: gnat-13
Version: 13.2.0-19
Severity: normal
X-Debbugs-Cc: lbre...@debian.org
Control: affects -1 pcscada libalog dbusada anet ahven libgmpada libgtkada
libgnatcoll-db libncursesada libaunit adacgi liblog4ada libtexttools
libtemplates-parser libxmlezout libgnatcoll-bindings libgnatcoll gp
Package: gnat-7
Version: 7.3.0-11
Severity: minor
Tags: upstream
Hello.
The following source
with Ada.Directories; use Ada.Directories;
with Ada.Text_IO; use Ada.Text_IO;
procedure Directories_Bug is
procedure Show (Path : in String) is
With_Slash : constant String := Path & '/';
.
Note: for historical reasons, gnatchop carries its own version
of the old algorithm, which seems to work in our case.
Bugs-Debian: https://bugs.debian.org/903694
Bugs-Debian: https://bugs.debian.org/856274
Author: Ludovic Brenta
Author: Nicolas Boulenguez
--- a/src/gcc/ada/osint.adb
+++
or historical reasons, gnatchop carries its own version
of the old algorithm, which seems to work in our case.
Bugs-Debian: https://bugs.debian.org/903694
Bugs-Debian: https://bugs.debian.org/856274
Author: Ludovic Brenta
Author: Nicolas Boulenguez
--- a/src/gcc/ada/osint.adb
+++ b/src/gcc/
gnat-7/7.3.0-26 wrongly assumes that /usr/bin/gcc links to gcc-7.
By chance, this currently holds in unstable,
and gcc-7 is hopefully able to build a fixed version of itself.
gnat-8/8.1.0-11 wrongly assumes that /usr/bin/gcc links to gcc-8.
It is broken in unstable, at least with default options.
Package: gnat-11
Version: 11.2.0-10
Severity: normal
Hello.
The following valid Ada sources crash 'gcc-11 -O2 -gnatn -gnatVa' as shown
below.
All goes well with gcc-10, with -O1, without inlining or without validity
checks.
cat > p.ads < p.adb < {heap 2040k} {heap 2040k}
{heap 2040k} {heap
/debian/patches/ada-lib-info-source-date-epoch.diff
+++ b/debian/patches/ada-lib-info-source-date-epoch.diff
@@ -62,24 +62,26 @@ Author: Nicolas Boulenguez
end File_Stamp;
function File_Stamp (Name : Path_Name_Type) return Time_Stamp_Type is
-@@ -3261,4 +3276,28 @@ begin
+@@ -3261,4
Source: gcc-13
Followup-For: Bug #1067453
Sorry for the poor summary. Here is what I have built and tested.
* native build on ppc46el (not amd64)
* gcc-source 13.2.0-19 from Debian
* libgnat-timet64.diff unapplied
(pr114065-proposed.diff should be unapplied for -20 and later)
* commits 1/8
Source: gcc-13
Followup-For: Bug #1067453
The gettimeofday import issue seems specific to the time_t 64
transition in Debian.
When building C on armhf, a #define replaces gettimeofday with
__gettimeofday64 so the linker finds the 64 bits version in the libc.
When linking Ada code, the linker sea
#1072328).
If Adrian can check that 0001, 0002 and 0003 fix #1072071,
then I suggest to merge them into a single commit,
else
if Matthias confirms that := was not deliberate
then I suggest to apply 0001.
From: Nicolas Boulenguez
Subject: [PATCH 1/3] probably fix a typo in e08cd8a2
--- a/debian
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 f
Package: gnat
Version: 4.3.2-3
Severity: minor
The following program crashes gnatmake. I could not make it
smaller. Removing the affectation (last line) or correcting the type
mismatch in the record declaration (first line) both causes a normal
error message.
Package: gnat-4.6
Version: 4.6.3-4
Severity: minor
Hello. Gnatmake calls gcc -shared in a way incompatible with
--as-needed. Gprbuild uses a better ordering.
-- bug.ads
package Bug is
end Bug;
-- bug.gpr
project Bug is
for Library_Name use "bug";
for Library_Version use "libbug.so.1";
for L
Package: ppl
Followup-For: Bug #670025
CREDITS.gz should not vary anymore with gzip >= 1.4-5,
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=647522
for the whole story.
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...
: Add libncursesw5-dev to Build-Depends due to
+missing library, fixes FTBFS. (Closes: #669479)
+
+ [ Nicolas Boulenguez ]
+ * Remove usr/bin/ppl_pl from libppl-swi multiarch package (see #670024).
+Document equivalent functionality with the shared library.
+ * control: libppl0.11-dev cannot be
Package: ppl
Version: 0.11.2-6.1
Severity: serious
Tags: patch pending
Justification: 0.11.2-6 did build on powerpc
Full build log:
https://buildd.debian.org/status/fetch.php?pkg=ppl&arch=powerpc&ver=0.11.2-6.1&stamp=1343423594
Last lines:
/bin/bash ../../libtool --tag=CXX --mode=link powerpc-
test suite with -gstabs rather than -g on powerpc, to avoid
+running out of memory, see Ubuntu bug #941676 (Closes: #683767)
+
+ -- Nicolas Boulenguez Fri, 03 Aug 2012 18:50:55 +0200
+
ppl (0.11.2-6.1) unstable; urgency=low
* Non-maintainer upload.
diff -Nru ppl-0.11.2/debian/rules ppl-0
retitle 579920 gnat-4.6: Assert_Failure sinfo.adb:2738
reassign 579920 gnat-4.6
found 579920 4.6.1-4
forwarded 579920 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50197
thanks
I produced the same failure with a smaller source.
$ gcc -c a.ads
+===GNAT BUG DETECTED===
reassign 276227 gnat-4.6
found 276227 4.6.1-5
thanks
Since both messages are now coherent, I propose to test one single type.
package Test_124 is
type T is range 1 .. 32;
type T_arr_unconstrained is array (T range <>) of boolean;
type T_arr_constrained is new T_arr_unconstrained (T);
reassign 278831 gnat-4.6
found 278831 4.6.1-5
thanks
I suggest providing a body to avoid any other illegality.
package Test_128 is
package inner is
private
type T1;
end inner;
type T1_ptr is access inner.T1; -- line 9 ERROR: gnat mistakenly accepts
end Test_128;
package body te
reassign 339356 gnat-4.6
found 4.6.1-5
thanks
The bug box now refers to
4.6.1 (x86_64-pc-linux-gnu) Assert_Failure atree.adb:794
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.d
reassign 427107 gnat-4.6
found 427107 4.6.1-5
thanks
We now have two distinct assertion failures.
First version with T1 Eq Neq gives:
4.6.1 (x86_64-pc-linux-gnu) Assert_Failure einfo.adb:904
Error detected at pak1.ads:1:1
Second version with T1 Eq and T2 Eq gives:
4.6.1 (x86_64-pc-linux-gnu) Ass
Actually, no illegal program is undetected. The (minor) bug is that GNAT
reports only one error instead of two.
If one line is commented out, GNAT correctly reports the problem for the
other one.
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". T
reassign 496614 gnat-4.6
retitle 496614 gnat: can not find definition of an ada variable with gdb
thanks
Some news from upstream.
4.1.2 buggy
4.1.3 ok
4.2.3 ok
4.3.1 buggy
4.6.1 ok
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact
retitle 575954 [fixed in 4.6] [amd64] GNAT.Expect.Non_Blocking_Spawn double
free or corruption
thanks
All works fine with gnat 4.6.1-5 on Linux 3.0.0-1-amd64.
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.deb
found 494945 4.4.6-2
thanks
| 4.4.6 (x86_64-pc-linux-gnu) Assert_Failure atree.adb:884 |
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110902211142
reassign 494945 gnat-4.6
found 494945 4.6.1-5
thanks
-- There were two distinct bugs.
package P is
type T1 is tagged null record;
type T2 is tagged null record;
function Func (Func_Formal : in T1'Class;
I : in Integer)
return access T2;
proce
The link provided by the submitter is now broken. It seems that
http://www.cs.mcgill.ca/~joerg/downloads/auction_system.tar.gz provides the
tarball that he describes.
With gnat-4.4-4.6-5 and libpolyorb1-dev-2.6.0~20090423-10 (providing
/usr/bin/po_gnatdist, emulation of the obsolete /usr/bin/gnatd
Gnat 4.6.1 does not crash on this source.
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110905231119.GA12906@pegase
Package: gnat-4.6
Version: 4.6.2-3
Severity: normal
Tags: upstream
Compiling the following program with gnatmake bug.adb displays
+===GNAT BUG DETECTED==+
| 4.6.2 (x86_64-pc-linux-gnu) Assert_Failure sinfo.adb:2947|
| Error detect
Package: gnat-4.6
Version: 4.6.2-3
Severity: normal
I reported the problem to
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52319).
The following legal program is rejected by "gnatgcc -c proc.adb" with
proc.adb:17:28: there is no applicable operator "And" for type
"Standard.Integer"
All works fin
Package: gnat-4.0
Version: 4.0.2-2
Severity: normal
Hello. A program failed to compile, and I tracked down the minimal
program causing the crash. I hope this will help finding the bug. The
source:
--
procedure Demo is
package P is
subtype T is Character;
Const : constant Characte
>Submitter-Id: net
>Originator:Nicolas Boulenguez
>Organization: none
>Confidential: no
>Synopsis: (Ada) Conversion of user-defined integer type with Size fixed
>causes crashes
>Severity: non-critical
>Priority: low
>Category: ada
>Cl
Package: gnat-4.9
Version: 4.9.1-4
Severity: minor
Tags: upstream
Hello.
The following source produces an error message.
generic_f.ads:2:48: "Result" attribute can only appear in postcondition of
function
It compiles when "generic" or the postcondition is removed.
generic
function Generic_F retu
Package: gnat-5
Followup-For: Bug #666106
With gnat-6/6.1.1-8, the following reproducer should display "No bug."
and causes a segmentation fault instead. This is most probably a new
symptom for the same issue.
# r.adb:
with Ada.Text_IO; use Ada.Text_IO;
procedure R is
F : File_Type;
begin
Package: src:gcc-6
Version: 6.1.1-10
Severity: minor
Tags: patch
Hello.
Please consider this trivial patch.
Description: updates for gnat-6 the script reproducing most Ada-related bugs.
Author: Nicolas Boulenguez
--- a/debian/ada/confirm_debian_bugs.py
+++ b/debian/ada/confirm_debian_bugs.py
Package: gnat-6
Followup-For: Bug #666106
Control: retitle -1 kfreebsd-i386, hurd-i386: Exceptions with tracebacks cause
STORAGE_ERROR or SEGFAULT
Adacontrol has been affected on kfreebds and hurd, see #833026.
Package: gnat-5
Version: 5.2.1-17
Severity: normal
Hello.
If main.adb contains:
procedure Main is begin null; end Main;
and is compiled with
# gcc-5 -c main.adb
# gnatbind-5 -x main.ali
# LANG=C gnatlink-5 main.ali -lgnatprj
the build fails:
/usr/lib/gcc/x86_64-linux-gnu/5/../../../x86_64-linux
Package: gnat-4.9
Followup-For: Bug #182360
Control: retitle -1 [Fixed in 5.2] GNAT.OS_Lib.Close ignores return value
Adacore has found a way to preserve compatibility *and* allow the user
to raise an exception.
>From s-os_lib.ads:
procedure Close (FD : File_Descriptor; Status : out Boolean);
Package: gnat-4.9
Followup-For: Bug #666106
Control: reassign -1 gnat-5 5.2.1-18
The message is "s-intman.adb:139 explicit raise" again.
Package: gnat-4.9
Followup-For: Bug #673772
Control: retitle: ATC with syscalls not working
Control: reassign -1 gnat-5 5.2.1-19
As I understand the reference manual, the following program
with Ada.Calendar; use Ada.Calendar;
with Ada.Text_IO; use Ada.Text_IO;
with Interfaces.C; use Interfa
Package: gnat-4.9
Followup-For: Bug #717014
Control: reassign -1 gnat-5 5.2.1-19
Confirmed on mips.
Package: gnat-4.9
Followup-For: Bug #760211
Control: reassign -1 gnat-5 5.2.1-21
Confirmed on kfreebsd-amd64.
Package: gnat-5
Followup-For: Bug #673772
Control: tag -1 upstream wontfix
Quoting upstream:
Actually Ada does not make any guarantee when mixing ATC and system calls and
indeed, it is fundamentally unsafe if not impossible to properly abort from any
random system call, so this isn't supported.
Package: gnat-5
Version: 5.2.1-22
Severity: wishlist
Hello.
The symbolic link
/usr/bin/gnatgcc -> gcc-x.y
installed by package gnat-x.y used to help mixing Ada and C code,
/usr/bin/gnatgcc version being compatible with gnat, and sometimes
different from /usr/bin/gcc version.
In gnat-5, the symli
Package: gnat-4.6
Version: 4.6.4-1
Severity: normal
On armel armhf ia64 m68k mips mipsel powerpc s390 s390x sparc, the
following scenario complains about the "cos", "sin", etc symbols being
undefined. Adding "-lm" to library_options (used at the end of the
link command) fixes the problem, but gnat
Package: gnat-4.6
Version: 4.6.4-1
Followup-For: Bug #687642
Control: retitle -1 gnat-4.6: on armel armhf, gcc -shared reads libgnarl-4.6.a
instead of .so
Hello. Here is a reproducer, tested on abel.debian.org (armel).
The bug may present other symptoms: ld will complain about missing
symbols, b
Package: gnat-4.6
Followup-For: Bug #687642
On amd64, /usr/lib/gcc/x86_64-linux-gnu/4.6/adalib/ contains
libgnarl.a
libgnarl.so -> ../../../../../x86_64-linux-gnu/libgnarl-4.6.so.1
On armhf, /usr/lib/gcc/arm-linux-gnueabi/4.6/adalib/ contains
libgnarl-4.6.a -> libgnarl.a
libgnarl.a
libgnarl.so ->
Followup-For: Bug #246187
Control: reassign -1 gnat-4.8 4.8.2-1
Control: retitle -1 Bug box, Storage_Error at system.ads:139:5 on legal program
4.8.2 (x86_64-linux-gnu) Storage_Error stack overflow or erroneous memory access
Error detected at system.ads:152:5
--
To UNSUBSCRIBE, email to debian-
Control: reassign -1 gnat-4.8 4.8.2-1
Control: retitle -1 Bug box on legal program in gnat_to_gnu_entity, at
ada/gcc-interface/decl.c
4.8.2 (x86_64-linux-gnu) GCC error:
in gnat_to_gnu_entity, at ada/gcc-interface/decl.c:582
Error detected at test_70.adb:18:9
--
To UNSUBSCRIBE, email to debian
Control: reassign -1 gnat-4.8 4.8.2-1
Control: retitle -1 Bug box in Case_Statement_to_gnu, at
ada/gcc-interface/trans.c:2198, on legal Ada 83 program
4.8.2 (x86_64-linux-gnu) GCC error
in Case_Statement_to_gnu, at ada/gcc-interface/trans.c:2198
Error detected at test_106.adb:4:9
--
To UNSUBSC
Control: retitle -1 [Fixed in 4.8] Assert_Failure atree.adb:794 on invalid code
(mixture of protected object and accept of entry family)
4.8.2-1 correctly detects the error:
gnat_bug.adb:13:10: enclosing body of accept must be a task
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.
Control: retitle -1 [Fixed in 4.8] Assert_Failure at atree.adb:886 caused by
legal prefixed notation
Both triggers produce the expected behaviour with 4.8.2-1:
gcc-4.8 -c trigger1.adb
gcc-4.8 -c p.ads
cannot generate code for file p.ads (package spec)
gnatmake: "p.ads" compilation error
--
To
Control: retitle -1 [Fixed in 4.8] Assert_Failure sinfo.adb:2738
4.8.2-1 produces the expected result:
a.ads:10:10: unmatched actual "Foo"
a.ads:10:10: in instantiation of "B" declared at line 5
--
To UNSUBSCRIBE, email to debian-gcc-requ...@lists.debian.org
with a subject of "unsubscribe". Tro
1 - 100 of 143 matches
Mail list logo
