Re: imagemagick@6.9.11-48 to graft or not to graft with 6.9.12-2

[Julien Lepiller]

I don't think I understand the problem fully, but it looks like there is nothing wrong with the graft now that you symlinked tge library, so it's fine to keep the graft. Hopefully we can ungraft shortly during the "ungraftathon" next week :)

Le 19 mars 2021 05:40:45 GMT-04:00, "Léo Le Bouter" <lle-bout@zaclys.net> a écrit :

Hello!

See commit: 82e887ba48c2ba91b17aa9b6b17501e3e0ef4aef

Following discussion around whether it is safe to graft and whether we
should do so or not, first, I apologize for not doing as rigorous
checking on this issue as I should have, and also requesting more peer-
review, I initially believed those two ImageMagick version were ABI
compatible with unchanged soname so it turns out it would be a rather
uncontroversial graft to make but now it turns out we have a changed
soname but whether it is binary (backwards) compatible or not remains a
question.

We had a user reporting that Inkscape stopped working after the graft (
https://logs.guix.gnu.org/guix/2021-03-18.log#100200), after which we
decided on IRC with rekado we might cheat by symlinking the shared
libraries, which I've done in commit
2e0ff59f0cd836b156f1ef2e78791d864ce3cfcd, from a glance it didnt seem
the soname change caused backwards incompatible changes but only
forward incompatible changes.

Let's see some abidiff output now:

$ ./pre-inst-env guix environment --ad-hoc libabigail -- abidiff
$(./pre-inst-env guix build --no-grafts imagemagick@6.9.11-48 | grep -v
doc)/lib/libMagickCore-6.Q16.so.6 $(./pre-inst-env guix build 
imagemagick@6.9.12-2g | grep -v doc)/lib/libMagickCore-6.Q16.so.7
ELF SONAME changed
Functions changes summary: 0 Removed, 0 Changed, 0 Added function
Variables changes summary: 0 Removed, 0 Changed, 0 Added variable
Function symbols changes summary: 0 Removed, 0 Added function symbol
not referenced by debug info
Variable symbols changes summary: 0 Removed, 1 Added variable symbol
not referenced by debug info

SONAME changed from 'libMagickCore-6.Q16.so.6' to 'libMagickCore-
6.Q16.so.7'

1 Added variable symbol not referenced by debug info:

  [A] .gomp_critical_user_analyzeImage


$ ./pre-inst-env guix environment --ad-hoc libabigail -- abidiff
$(./pre-inst-env guix build --no-grafts imagemagick@6.9.11-48 | grep -v
doc)/lib/libMagick++-6.Q16.so.8 $(./pre-inst-env guix build 
imagemagick@6.9.12-2g | grep -v doc)/lib/libMagick++-6.Q16.so.9
ELF SONAME changed
Functions changes summary: 0 Removed, 0 Changed, 0 Added function
Variables changes summary: 0 Removed, 0 Changed, 0 Added variable

SONAME changed from 'libMagick++-6.Q16.so.8' to 'libMagick++-
6.Q16.so.9'

$ ./pre-inst-env guix environment --ad-hoc libabigail -- abidiff
$(./pre-inst-env guix build --no-grafts imagemagick@6.9.11-48 | grep -v
doc)/lib/libMagickWand-6.Q16.so.6 $(./pre-inst-env guix build 
imagemagick@6.9.12-2g | grep -v doc)/lib/libMagickWand-6.Q16.so.7
ELF SONAME changed
Functions changes summary: 0 Removed, 0 Changed, 0 Added function
Variables changes summary: 0 Removed, 0 Changed, 0 Added variable

SONAME changed from 'libMagickWand-6.Q16.so.6' to 'libMagickWand-
6.Q16.so.7'

Any more ABI diff-ing/testing, information, etc.. on whether this is
safe or not is welcome, it sounds to me it could be fine but there is
some amount of doubt still.

If we can't graft ImageMagick we shall revert all commits and then it
means we would have to apply patches for each and every CVE which can
be tedious to create and maintain and to me leaving the package as-is
without patching is not really OK :-/

To graft or not to graft?

Thank you,
Léo