Bug#1063719: More analysis and improved patch

2024-02-13 Thread George Robbert
Hey, Just like with the previous bug (#1026927), it looks like there's more to this one. Trying the patch on several more systems I run into the same symptoms on some. Again, these have the same symptoms but differing causes (full patch included). What I found was: On an intel system, where

Bug#1063719: needrestart -b complainst of initialized variables on some AMD and all ARM

2024-02-11 Thread George Robbert
Package: needrestart Version: 3.6-4+deb12u1 Severity: normal Tags: patch Dear Maintainer, When running 'needrestart -b' on some AMD systems I get the following uninitialized variable warning. It also does not report the expected microcode version (NEEDRESTART-UCEXP). Use of uninitialized value

Bug#1026927: More analysis and improved patch

2023-01-14 Thread George Robbert
Hey, I think I found why you still get the error after applying the patch. The same thing happened to me on another of my systems. The difference was whether the package amd64-microcode was installed or not. I think I have a patch (attached) for both. Here's what I found. It looks like

Bug#1026927: needrestart -b on AMD processors complains of perl uninitialized variable

2023-01-09 Thread George Robbert
On Fri, 06 Jan 2023 01:56:33 -0700, Patrick Matthäi wrote: > Thanks for your investigation. I have tested it on my (sid/unstable) AMD > notebook and I can reproduce your issue with needrestart from stable. > But your patch does not produce another result for me. Are you sure, > that you didnt

Bug#1026927: needrestart -b on AMD processors complains of perl uninitialized variable

2022-12-23 Thread George Robbert
eedRestart/uCode/AMD.pm ends in a comma (,) instead of a semicolon (;). This means that assignment is subsumed into the next line which is under if ($debug). See the attached patch. It also looks to me like this is similar symptoms with a different root cause from bug #973050. Thanks, Geor

Bug#958353: sleepd: fails to correctly read acpi subsystem version

2020-04-20 Thread George Robbert
Package: sleepd Version: 2.10 Severity: normal Tags: patch Dear Maintainer, When running with a newer kernel (4.19.0-8-amd64 from buster, though this also occurs with 5.4.0-0.bpo.4-amd64 from backports), and without upower installed, sleepd fails to start, complaining ACPI subsystem