| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20191105045239.GC27350@Gentoo>
Date: Tue, 5 Nov 2019 10:22:42 +0530
From: Bhaskar Chowdhury <unixbhaskar@...il.com>
To: "J. Bruce Fields" <bfields@...ldses.org>
Cc: Randy Dunlap <rdunlap@...radead.org>,
yamada.masahiro@...ionext.com, michal.lkml@...kovi.net,
linux-kbuild@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] scripts:prune-kernel:remove old kernels and modules dir
from system
On 10:09 Tue 05 Nov 2019, Bhaskar Chowdhury wrote:
>On 21:32 Mon 04 Nov 2019, J. Bruce Fields wrote:
>>On Mon, Nov 04, 2019 at 06:03:13PM -0800, Randy Dunlap wrote:
>>> On 11/1/19 11:30 PM, Bhaskar Chowdhury wrote:
>>> > This patch allow you to remove old kernels and associated modules
>>> > directory from the system.You can do it at once with the -r flag
>>> > and interactively with the -i flag.
>>> >
>>> > Signed-off-by: Bhaskar Chowdhury <unixbhaskar@...il.com>
>>> > ---
>>> > scripts/prune-kernel | 82 +++++++++++++++++++++++++++++++++++---------
>>> > 1 file changed, 65 insertions(+), 17 deletions(-)
>>>
>>> Hi,
>>> I believe that this script now does what the patch author intends it to do.
>>> It does have a few whitespace issues, but no big deals. (see below)
>>
>>My original comment stands: looks like it prompts for full module path
>>and kernel versions which means it's no more convenient than just doing
>>an "ls" and then removing the ones you want to. (In fact, with "rm"
>>you'd also get the benefit of tab completion....)
Bruce,
There is -r or --remove option , with that you can pass the kernel
version and modules directory name , and the script will remove those
without prompting you for those.
For instance you need to pass the options like this :
./scripts/prune-kernel -r kernel_version modules_directory_name
OR
./scripts/prune-kernel --remove kernel_version modules_directory_name
Which I certainly believe is not just "ls and rm" stuff...could you
please give it a shot ,and let me know the shortcoming please??
So , I can fix the damn thing.
>
>Hi Bruce,
>I am forcing user to put full modules directory name,kinda engagement
>to the process.It is not a trivial operation, so the users should be
>aware.
>OHT Could you please state what else you would like to be inducted in
>the so it can be useful?? So sorry for my weak memory if you already
>explicitly stated something like that before.
>
>Thanks,
>Bhaskar
>>
>>It's quite different from the original script and I don't really see the
>>advantage.
>>
>>--b.
>>
>>>
>>> Tested-by: Randy Dunlap <rdunlap@...radead.org>
>>>
>>>
>>> > diff --git a/scripts/prune-kernel b/scripts/prune-kernel
>>> > index e8aa940bc0a9..01d0778db71f 100755
>>> > --- a/scripts/prune-kernel
>>> > +++ b/scripts/prune-kernel
>>> > @@ -1,21 +1,69 @@
>>> > #!/bin/bash
>>> > # SPDX-License-Identifier: GPL-2.0
>>> > +#This script will remove old kernels and modules directory related to it.
>>> > +#"-r" or "--remove" show how to silently remove old kernel and modules dir.
>>> > +# "-h" or "--help" show how to use this script or show without parameter.
>>> > +#"-i" or "--interactive" show how to remove interactively.
>>> > +
>>> > +flag=$1
>>> > +kernel_version=$2
>>> > +modules_version=$3
>>> > +boot_dir=/boot
>>> > +modules_dir=/lib/modules
>>> > +
>>> > +remove_old_kernel() {
>>> > + cd $boot_dir
>>> > + rm -If vmlinuz-$kernel_version System.map-$kernel_version config-$kernel_version
>>> > + return 0
>>> > +}
>>> > +
>>> > +remove_old_modules_dir() {
>>> > + cd $modules_dir
>>> > + rm -rf $modules_version
>>> > + return 0
>>> > +}
>>> > +
>>> > +usage() {
>>> > + printf "Usage: $(basename $0) [-ri]\n"
>>> > + printf "\n -r or --remove kernel_version modules_version\n"
>>> > + printf "\n -i or --interactive do as interactive way\n"
>>> > + return 0
>>> > +}
>>> > +
>>> > +case "$flag" in
>>> > + -i | --interactive)
>>> > + printf "\nEnter kernel version to remove or blank/empty to exit:"
>>> > + read kernel_version
>>> > + if [[ $kernel_version != "" ]]; then
>>> > + remove_old_kernel
>>> > + printf "\nRemoved kernel version:$kernel_version from the system.\n\n"
>>>
>>> space after ':'
>>>
>>> drop one \n above.
>>>
>>> > + printf "Please give the full modules directory name to remove:"
>>> > + read modules_version
>>> > + if [[ $modules_version != "" ]]; then
>>> > + remove_old_modules_dir
>>> > + printf "\n\nRemoved modules directory:$modules_version from the system.\n\n"
>>>
>>> space after ':'
>>>
>>> drop one \n above.
>>>
>>> > + else
>>> > + exit 1
>>> > + fi
>>> > + fi
>>> > + ;;
>>> > + -h | --help)
>>> > + usage
>>> > + exit 0
>>> > + ;;
>>> > + -r | --remove)
>>> > + if [[ $# -ne 3 ]]; then
>>> > + printf "You need to provide kernel version and modules directory name.\n"
>>> > + exit 1
>>> > + else
>>> > + remove_old_kernel
>>> > + remove_old_modules_dir
>>> > + fi
>>> > + ;;
>>> > + *)
>>> > + usage
>>> > + exit 1
>>> > + ;;
>>> > +esac
>>> >
>>> > -# because I use CONFIG_LOCALVERSION_AUTO, not the same version again and
>>> > -# again, /boot and /lib/modules/ eventually fill up.
>>> > -# Dumb script to purge that stuff:
>>> >
>>>
>>> OK, the former script's loop is removed.. good.
>>> But the 2 preceding blank lines are not removed, so the script
>>> now ends with 2 unnecessary blank lines.
>>>
>>> > -for f in "$@"
>>> > -do
>>> > - if rpm -qf "/lib/modules/$f" >/dev/null; then
>>> > - echo "keeping $f (installed from rpm)"
>>> > - elif [ $(uname -r) = "$f" ]; then
>>> > - echo "keeping $f (running kernel) "
>>> > - else
>>> > - echo "removing $f"
>>> > - rm -f "/boot/initramfs-$f.img" "/boot/System.map-$f"
>>> > - rm -f "/boot/vmlinuz-$f" "/boot/config-$f"
>>> > - rm -rf "/lib/modules/$f"
>>> > - new-kernel-pkg --remove $f
>>> > - fi
>>> > -done
>>> > --
>>>
>>>
>>> --
>>> ~Randy
>-----BEGIN PGP SIGNATURE-----
>
>iQEzBAABCAAdFiEEnwF+nWawchZUPOuwsjqdtxFLKRUFAl3A/PMACgkQsjqdtxFL
>KRUyqQgAya3BYL2iEMk0N3b1/Ji3TVzjrhaMrYM1bc6MDs7UA83ikBBG7C/dSmZ9
>3l9ANXfEnjaF64FpuLTNE2pT/tVxEj+xQz5iWw3JXV9TR7P2rq6814LH4b3hXX54
>A9r0sIRz0xwCMq7ihtNtgXziyviLp8UkUjIK+1C4vQ0h9/CvP0YETq8fgRWASMrS
>2+VErRPd06DryAAnapz9CnZz2Fju6S4mA2aaTJ3hiDLNmbj++6IoNQw+r5MZrWde
>yUqPQcfAzxs6/GIuUPK3uStudgofQjCKbHbPho9jjmwXrNSFH6/0/pjz5R91oMu9
>AWpkl0vkvAvzLsjRIeMG/MmClx0c3w==
>=VReA
>-----END PGP SIGNATURE-----
Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)
Powered by blists - more mailing lists