Debian Bug report logs - #651402
iulib: FTBFS: Checking for inflate() in C library tiff... no

Toggle useless messages

Message #3 received at submit@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: submit@bugs.debian.org

Subject: iulib: FTBFS: Checking for inflate() in C library tiff... no

Date: Thu, 8 Dec 2011 11:58:52 +0100

Source: iulib
Version: 0.4.4+ds-1
Severity: serious
Justification: fails to build from source

iulib in experimental FTBFS:
|    debian/rules override_dh_auto_build
| make[1]: Entering directory `/build/buildd-iulib_0.4.4+ds-1-s390x-VaiIvL/iulib-0.4.4+ds'
| scons -j5 prefix=/usr soversion=0d
| scons: Reading SConscript files ...
| compiling with high optimization
| Checking for png_byte in C library png... yes
| Checking for jpeg_std_error() in C library jpeg... yes
| Checking for inflate() in C library tiff... no
| AssertionError: please install: libtiff4-dev:
|   File "/build/buildd-iulib_0.4.4+ds-1-s390x-VaiIvL/iulib-0.4.4+ds/SConstruct", line 65:
|     assert conf.CheckLibWithHeader('tiff', 'tiff.h', 'C', 'inflate();', 1), "please install: libtiff4-dev"
| make[1]: *** [override_dh_auto_build] Error 2
| make[1]: Leaving directory `/build/buildd-iulib_0.4.4+ds-1-s390x-VaiIvL/iulib-0.4.4+ds'
| make: *** [build] Error 2

Full build log:
https://buildd.debian.org/status/fetch.php?pkg=iulib&arch=s390x&ver=0.4.4%2Bds-1&stamp=1323313412
(but I can reproduce it on i386)

-- 
Jakub Wilk

Message #8 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Bernhard Reiter <ockham@raz.or.at>

To: 651402@bugs.debian.org

Subject: iulib: FTBFS: Checking for inflate() in C library tiff... no

Date: Wed, 15 Feb 2012 23:11:44 +0100

Upstream's bugtracker seems to have a fix:

http://code.google.com/p/iulib/issues/detail?id=27#c2

Message #15 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jeffrey Ratcliffe <jjr@debian.org>

To: 651402@bugs.debian.org

Subject: Symbols in iulib

Date: Sun, 1 Apr 2012 14:20:48 +0200

The suggested patch indeed fixes the problem. However, the symbols
exported are now updated. I assume this means that the SONAME should
be bumped.

To what? 0e? How was 0d chosen? d for debian?

The packaging for ocropus should then be updated to fix the dependency
on libiulib0d.

Message #18 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Sun, 1 Apr 2012 22:34:19 +0200

* Jeffrey Ratcliffe <jjr@debian.org>, 2012-04-01, 14:20:
>The suggested patch indeed fixes the problem. However, the symbols 
>exported are now updated.

What do you mean?

>I assume this means that the SONAME should be bumped.

The patch (or mere recompilation) should not break ABI, unless something 
is very very broken.

>To what? 0e? How was 0d chosen? d for debian?

Yes, "d" stands for Debian.

-- 
Jakub Wilk

Message #23 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jeffrey Ratcliffe <jeffrey.ratcliffe@gmail.com>

To: Jakub Wilk <jwilk@debian.org>, 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Mon, 2 Apr 2012 21:52:25 +0200

On 1 April 2012 22:34, Jakub Wilk <jwilk@debian.org> wrote:
>> The suggested patch indeed fixes the problem. However, the symbols
>> exported are now updated.
>
> What do you mean?

I get a diff that starts like this, and carries on for a couple of
hundred lines:

dpkg-gensymbols: warning: some new symbols appeared in the symbols
file: see diff output below
dpkg-gensymbols: warning: some symbols or patterns disappeared in the
symbols file: see diff output below
dpkg-gensymbols: warning: debian/libiulib0d/DEBIAN/symbols doesn't
match completely debian/libiulib0d.symbols
--- debian/libiulib0d.symbols (libiulib0d_0.4.4+ds-2_i386)
+++ dpkg-gensymbols9n1RFj	2012-03-31 21:36:53.332455360 +0200
@@ -2,53 +2,67 @@
  _ZN14SDL_extensions13SDL_draw_lineEP11SDL_Surfaceiiiij@Base 0.4.4
  (optional)_ZN5colib4copyIN6imgrle6RLERunES2_EEvRNS_6narrayIT_EERKNS3_IT0_EE@Base
0.4.4
  (optional)_ZN5colib4copyIffEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
- (optional)_ZN5colib4copyIhfEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
- (optional)_ZN5colib4copyIhhEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
- (optional)_ZN5colib4copyIiiEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
- (optional)_ZN5colib4moveIfEEvRNS_6narrayIT_EES4_@Base 0.4.4
- (optional)_ZN5colib4moveIhEEvRNS_6narrayIT_EES4_@Base 0.4.4
- (optional)_ZN5colib4moveIiEEvRNS_6narrayIT_EES4_@Base 0.4.4
- _ZN5colib5stdioD1Ev@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4copyIhfEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4copyIhhEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4copyIiiEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4moveIfEEvRNS_6narrayIT_EES4_@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4moveIhEEvRNS_6narrayIT_EES4_@Base 0.4.4
+#MISSING: 0.4.4+ds-2#
(optional)_ZN5colib4moveIiEEvRNS_6narrayIT_EES4_@Base 0.4.4
+#MISSING: 0.4.4+ds-2# _ZN5colib5stdioD1Ev@Base 0.4.4

Message #26 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Mon, 2 Apr 2012 23:06:42 +0200

* Jeffrey Ratcliffe <jeffrey.ratcliffe@gmail.com>, 2012-04-02, 21:52:
>--- debian/libiulib0d.symbols (libiulib0d_0.4.4+ds-2_i386)
>+++ dpkg-gensymbols9n1RFj	2012-03-31 21:36:53.332455360 +0200
>@@ -2,53 +2,67 @@
>  _ZN14SDL_extensions13SDL_draw_lineEP11SDL_Surfaceiiiij@Base 0.4.4
>  (optional)_ZN5colib4copyIN6imgrle6RLERunES2_EEvRNS_6narrayIT_EERKNS3_IT0_EE@Base
>0.4.4
>  (optional)_ZN5colib4copyIffEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>- (optional)_ZN5colib4copyIhfEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>- (optional)_ZN5colib4copyIhhEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>- (optional)_ZN5colib4copyIiiEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>- (optional)_ZN5colib4moveIfEEvRNS_6narrayIT_EES4_@Base 0.4.4
>- (optional)_ZN5colib4moveIhEEvRNS_6narrayIT_EES4_@Base 0.4.4
>- (optional)_ZN5colib4moveIiEEvRNS_6narrayIT_EES4_@Base 0.4.4
>- _ZN5colib5stdioD1Ev@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4copyIhfEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4copyIhhEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4copyIiiEEvRNS_6narrayIT_EERKNS1_IT0_EE@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4moveIfEEvRNS_6narrayIT_EES4_@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4moveIhEEvRNS_6narrayIT_EES4_@Base 0.4.4
>+#MISSING: 0.4.4+ds-2#
>(optional)_ZN5colib4moveIiEEvRNS_6narrayIT_EES4_@Base 0.4.4
>+#MISSING: 0.4.4+ds-2# _ZN5colib5stdioD1Ev@Base 0.4.4

This part looks innocent. :)

Could you include the full diff? (I'm sorry, I don't have time to try to 
build the package myself.)

-- 
Jakub Wilk

Message #31 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jeffrey Ratcliffe <jjr@debian.org>

To: Jakub Wilk <jwilk@debian.org>, 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Tue, 3 Apr 2012 20:58:13 +0200

[Message part 1 (text/plain, inline)]

I've attached the build log.

[iulib_0.4.4+ds-2_i386.build (application/octet-stream, attachment)]

Message #34 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Wed, 4 Apr 2012 21:17:50 +0200

AFAICS all the symbols that disappeared correspond to functions declared 
as inline. According to 
http://lists.debian.org/debian-devel/2012/01/msg00755.html it's safe to 
mark such symbols as optional.

-- 
Jakub Wilk

Message #39 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jeffrey Ratcliffe <jeffrey.ratcliffe@gmail.com>

To: Jakub Wilk <jwilk@debian.org>, 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Sun, 15 Apr 2012 11:58:44 +0200

You seem to be distinguishing between functions declared as inline,
and those inlined by the compiler. My understanding that the compiler
was under no obligation to follow the inline directive, and I don't
see an easy way to check when functions have been inlined or not.

Taking an example - this symbol has disappeared and is not marked as optional:

- _ZN5iulib11ChainTracer9nextpixelEv@Base 0.4.4

It is a function in a struct. I assume it should be optional.

Can you explain?

Message #42 received at 651402@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: 651402@bugs.debian.org

Subject: Re: Bug#651402: Symbols in iulib

Date: Mon, 16 Apr 2012 13:44:57 +0200

* Jeffrey Ratcliffe <jeffrey.ratcliffe@gmail.com>, 2012-04-15, 11:58:
>You seem to be distinguishing between functions declared as inline, and 
>those inlined by the compiler. My understanding that the compiler was 
>under no obligation to follow the inline directive, and I don't see an 
>easy way to check when functions have been inlined or not.

What really matters is that a function is _declared_ inline or not. 
Symbols for functions declared inline could dissapear at any time 
without constituting ABI breakage (AFAIUI).

>Taking an example - this symbol has disappeared and is not marked as 
>optional:
>
>- _ZN5iulib11ChainTracer9nextpixelEv@Base 0.4.4
>
>It is a function in a struct. I assume it should be optional.

A member function that is declared within the class body is considered 
as declared inline.[0] So yes, this one should be marked optional.

>Can you explain?

I hope it makes things clear.

But anyway, if you don't beat me to it, I'll tidy up iulib myself on the 
next weekend.

[0] http://www.parashift.com/c++-faq-lite/inline-functions.html#faq-9.8

-- 
Jakub Wilk

Message #49 received at 651402-close@bugs.debian.org (full text, mbox, reply):

From: Jakub Wilk <jwilk@debian.org>

To: 651402-close@bugs.debian.org

Subject: Bug#651402: fixed in iulib 0.4.4+ds-2

Date: Sat, 21 Apr 2012 21:18:04 +0000

Source: iulib
Source-Version: 0.4.4+ds-2

We believe that the bug you reported is fixed in the latest version of
iulib, which is due to be installed in the Debian FTP archive:

iulib_0.4.4+ds-2.debian.tar.gz
  to main/i/iulib/iulib_0.4.4+ds-2.debian.tar.gz
iulib_0.4.4+ds-2.dsc
  to main/i/iulib/iulib_0.4.4+ds-2.dsc
libiulib-dev_0.4.4+ds-2_i386.deb
  to main/i/iulib/libiulib-dev_0.4.4+ds-2_i386.deb
libiulib0d_0.4.4+ds-2_i386.deb
  to main/i/iulib/libiulib0d_0.4.4+ds-2_i386.deb



A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 651402@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Jakub Wilk <jwilk@debian.org> (supplier of updated iulib package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Format: 1.8
Date: Sat, 21 Apr 2012 22:46:25 +0200
Source: iulib
Binary: libiulib0d libiulib-dev
Architecture: source i386
Version: 0.4.4+ds-2
Distribution: experimental
Urgency: low
Maintainer: Jeffrey Ratcliffe <Jeffrey.Ratcliffe@gmail.com>
Changed-By: Jakub Wilk <jwilk@debian.org>
Description: 
 libiulib-dev - C++ library of image understanding-related algorithms -- developm
 libiulib0d - C++ library of image understanding-related algorithms
Closes: 651402 662381
Changes: 
 iulib (0.4.4+ds-2) experimental; urgency=low
 .
   * Update the watch file.
   * Remove the symbol file. It's proven to be maintenance burden and doesn't
     provide significant benefits, as each upstream release is likely to be
     binary incompatible anyway. Add lintian override for
     no-symbols-control-file.
   * Use TIFFOpen to check for libtiff existence. This fixed FTBFS with
     --no-add-needed (closes: #651402).
   * Remove unused build-dependency on imagemagick.
   * Bump debhelper compatibility level to 8 (no changes needed).
   * Multi-archify:
     + Bump minimum required debhelper version to 8.1.3.
     + Make libiulib0d and libulib-dev ‘Multi-Arch: same’.
     + Add ‘Pre-Depends: ${misc:Pre-Depends}’ to libiulib0.
     + Install shared libraries to /usr/lib/$(DEB_HOST_MULTIARCH) rather than
       /usr/lib.
     + Bump minimum required version of dpkg-dev to 1.16.
   * Bump standards version to 3.9.3 (no changes needed).
   * Drop DM-Upload-Allowed field, no longer needed.
   * Honour {CXX,CPP,LD}FLAGS from dpkg-buildflags.
   * Build-depend on libpng-dev instead of libpng12-dev (closes: #662381).
     Thanks to Nobuhiro Iwamatsu for the bug report.
   * Don't ignore errors from scons in the clean target.
Checksums-Sha1: 
 6cbed0790355e20b8e7f8e15e4a86932bb2cf5ec 1924 iulib_0.4.4+ds-2.dsc
 bfe54d3f544d37ccde9ff509a2a1a9202b15f8a9 5786 iulib_0.4.4+ds-2.debian.tar.gz
 8c7f5cb6b4cdfc91e9662e751b59b2d8850d60e1 204364 libiulib0d_0.4.4+ds-2_i386.deb
 abef288e0575e5cd0764b71fbedb2740fee5544c 61046 libiulib-dev_0.4.4+ds-2_i386.deb
Checksums-Sha256: 
 a6eaa8ea8f81aec6b349e05e26405bf3b6d5fec9d47485da1e6f6a34c2254063 1924 iulib_0.4.4+ds-2.dsc
 e9e6434372d438db169b2fb63b96a3eb22c403e79ca7de589452a6218c6b0812 5786 iulib_0.4.4+ds-2.debian.tar.gz
 fd7b843a1ca058bacc72cbef7bf7b5f55ceacbb070673d92bf1f515d8341066c 204364 libiulib0d_0.4.4+ds-2_i386.deb
 6bd0de635e392bb1233b0cada0c6afb455539241461015a0354d0d872014e0ca 61046 libiulib-dev_0.4.4+ds-2_i386.deb
Files: 
 44f6a0908f2b5237b3032c20e63dfa69 1924 graphics optional iulib_0.4.4+ds-2.dsc
 eb4b9f311e48b34448ce760339e7488b 5786 graphics optional iulib_0.4.4+ds-2.debian.tar.gz
 1658512e6105c592e0bdb1628723f2da 204364 libs optional libiulib0d_0.4.4+ds-2_i386.deb
 6ed88ec5e9bb0deedbd04809d3bf1162 61046 libdevel optional libiulib-dev_0.4.4+ds-2_i386.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
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=X4H8
-----END PGP SIGNATURE-----

Send a report that this bug log contains spam.