apt, branch 2.7.10Debians commandline package managerhttps://git.kalnischkies.de/apt/atom?h=2.7.102024-01-19T08:33:31ZRelease 2.7.102024-01-19T08:33:31ZJulian Andres Klodejulian.klode@canonical.com2024-01-19T08:31:50Zurn:sha1:3663365969a496e590bc70f448dbdf0bec52f2a2Add Conflicts: apt-verify2024-01-19T08:30:39ZJulian Andres Klodejulian.klode@canonical.com2024-01-19T08:25:31Zurn:sha1:910924ed7ecacc538e082ff9a81914119ba7838c
The APT team is very concerned about the encroachement of its
namespace and the impact on security of its file verification
process. We have expressed those concerns in the ITP bug, but
the package was nonetheless uploaded and accepted, so we have
to take this extraordinary step to protect our users.
Gbp-Dch: full
Merge branch 'happy-valgrind' into 'main'2024-01-12T14:22:47ZJulian Andres Klodejak@debian.org2024-01-12T14:22:47Zurn:sha1:8dc2b38047ee5db85166075d8292e1932ea64193
pkgcachegen: Use placement new to construct header
See merge request apt-team/apt!320pkgcachegen: Use placement new to construct header2024-01-12T12:49:24ZJulian Andres Klodejulian.klode@canonical.com2024-01-12T12:47:52Zurn:sha1:0ee745995b1f45ed321b0c20b151ec8e76e4344f
Avoid copying the header from a stack allocated object as this
will copy uninitialized padding bytes into the cache, triggering
valgrind errors which people then use as a strawman for unrelated
errors on armhf.
In an optimal world we should annotate the allocator however such
that valgrind actually does treat those bytes as uninitialized and
then supress warnings in the harmless places, such that when you
then go and try to access it in a place that matters, you do get
an error for uninitialized memory.
Currently any access within the pool will be considered initialized
which is clearly suboptimal. But this is very much a TBD topic and
involves annotating the allocator everywhere.
Release 2.7.92024-01-11T09:06:14ZJulian Andres Klodejulian.klode@canonical.com2024-01-11T09:05:29Zurn:sha1:ed053de346d92c1a4218a7b80a20158f3535fe2aCI: Pull from testing, unstable broken atm2024-01-10T12:32:24ZJulian Andres Klodejulian.klode@canonical.com2024-01-10T12:32:24Zurn:sha1:ecd2081a9aab7744771c896b5d7db2f70f120e17Accept file system disorder in test-ignored-files2024-01-10T11:06:23ZDavid Kalnischkiesdavid@kalnischkies.de2024-01-10T10:48:51Zurn:sha1:d5cdd79bfb1b0dd1cadd878dad8273cbf7069dbb
Reading the contents of a directory is not deterministic, so if we
wanted a fixed order we would need to sort the reported errors, but
as we don't need any specific order lets just accept both possibilities.
Regression-of: 7b41275b9da31d6c87bbaa0c9115e224e47b15e1
Release 2.7.82024-01-08T10:46:15ZJulian Andres Klodejulian.klode@canonical.com2024-01-08T10:46:15Zurn:sha1:559ad32ed4d88e3216c3e12197bb54a9364f733fRevert "Merge branch 'distclean-doc-an-test' into 'main'"2024-01-08T09:40:10ZJulian Andres Klodejulian.klode@canonical.com2024-01-08T09:39:46Zurn:sha1:10367ebb0f44cbc61b48fb99e194301c2182fe31
This reverts commit 86e6eace1d50527b5a2396290acd1db819b13e26, reversing
changes made to 6e43eef9ca8250eb561f2c9af2f4890d674f3911.
test: Disable valgrind on armhf, incompatible with stack clash protector2024-01-08T09:19:02ZJulian Andres Klodejulian.klode@canonical.com2024-01-08T09:12:42Zurn:sha1:719676d84b4bcdc5c65b87d42c4b72812209832b
Closes: #1059352