[med-svn] [Git][med-team/gmap][master] 6 commits: d/rules: enforce a given simd level for build reproducibility.

Étienne Mollier (@emollier) gitlab at salsa.debian.org
Sun Oct 2 17:08:01 BST 2022



Étienne Mollier pushed to branch master at Debian Med / gmap


Commits:
d92b1168 by Étienne Mollier at 2022-10-01T22:32:44+02:00
d/rules: enforce a given simd level for build reproducibility.

- - - - -
e26acec1 by Étienne Mollier at 2022-10-02T17:43:35+02:00
d/control: add myself to uploaders.

- - - - -
0f26d2e5 by Étienne Mollier at 2022-10-02T17:48:52+02:00
d/rules: disable simd on non-amd64 and limit to sse2 on amd64.

Closes: #1020793

- - - - -
e212d840 by Étienne Mollier at 2022-10-02T17:58:29+02:00
d/gmap.lintian-overrides: update mismatched override.

- - - - -
31c08548 by Étienne Mollier at 2022-10-02T17:59:03+02:00
d/control: declare compliance to standards version 4.6.1.

- - - - -
57046183 by Étienne Mollier at 2022-10-02T18:00:45+02:00
ready to upload to unstable.

- - - - -


4 changed files:

- debian/changelog
- debian/control
- debian/gmap.lintian-overrides
- debian/rules


Changes:

=====================================
debian/changelog
=====================================
@@ -1,3 +1,18 @@
+gmap (2021-12-17+ds-2) unstable; urgency=medium
+
+  [ Nilesh Patra ]
+  * [skip ci] Remove myself from uploaders
+
+  [ Étienne Mollier ]
+  * d/rules: enforce a given simd level for build reproducibility.
+  * d/control: add myself to uploaders.
+  * d/rules: disable simd on non-amd64 and limit to sse2 on amd64.
+    (Closes: #1020793)
+  * d/gmap.lintian-overrides: update mismatched override.
+  * d/control: declare compliance to standards version 4.6.1.
+
+ -- Étienne Mollier <emollier at debian.org>  Sun, 02 Oct 2022 18:00:07 +0200
+
 gmap (2021-12-17+ds-1) unstable; urgency=medium
 
   * New upstream version


=====================================
debian/control
=====================================
@@ -3,6 +3,7 @@ Maintainer: Debian Med Packaging Team <debian-med-packaging at lists.alioth.debian.
 Uploaders: Shaun Jackman <sjackman at debian.org>,
            Andreas Tille <tille at debian.org>,
            Alexandre Mestiashvili <mestia at debian.org>,
+           Étienne Mollier <emollier at debian.org>
 Section: science
 Priority: optional
 Build-Depends: debhelper-compat (= 13),
@@ -10,7 +11,7 @@ Build-Depends: debhelper-compat (= 13),
                libbz2-dev,
                zlib1g-dev,
                debhelper
-Standards-Version: 4.6.0
+Standards-Version: 4.6.1
 Vcs-Browser: https://salsa.debian.org/med-team/gmap
 Vcs-Git: https://salsa.debian.org/med-team/gmap.git
 Homepage: http://research-pub.gene.com/gmap


=====================================
debian/gmap.lintian-overrides
=====================================
@@ -1 +1 @@
-gmap: spelling-error-in-binary usr/lib/gmap/gmap* ofO of
+gmap: spelling-error-in-binary ofO of [usr/lib/gmap/*]


=====================================
debian/rules
=====================================
@@ -10,7 +10,15 @@ include /usr/share/dpkg/architecture.mk
 
 HELP2MAN = /usr/bin/help2man --no-info --version-string=${DEB_VERSION}
 
-ifneq (,$(filter $(DEB_HOST_ARCH), i386))
+# FIXME: gmap seems to have all the cpuid control necessary to avoid baseline
+# violation when compiled with high simd level (e.g. avx512bw) but running on
+# low baseline CPU (e.g. sse4_2).  The problem is that the build fails when
+# attempting to compile high baseline on a build system with low baseline.  It
+# would be nice to be able to provide high simd level to high baseline CPU as
+# well without build failure.  See #1020793 for more details.
+ifeq (,$(filter $(DEB_HOST_ARCH), amd64))
+  SIMD = --enable-simd --with-simd-level=sse2
+else
   SIMD = --disable-simd
 endif
 



View it on GitLab: https://salsa.debian.org/med-team/gmap/-/compare/5e46c7ba9a701bb2e10363000174507a70e7d6f7...57046183fd1e936f0f29864bdd75e19065487531

-- 
View it on GitLab: https://salsa.debian.org/med-team/gmap/-/compare/5e46c7ba9a701bb2e10363000174507a70e7d6f7...57046183fd1e936f0f29864bdd75e19065487531
You're receiving this email because of your account on salsa.debian.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/debian-med-commit/attachments/20221002/0bbe786b/attachment-0001.htm>


More information about the debian-med-commit mailing list