linux-kernel - Re: [PATCH] scsi:iscsi: Record session's startup mode in kernel

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <717da158-5a7e-b478-61d3-3753b0b00e01@oracle.com>
Date:   Wed, 23 Nov 2022 11:21:42 -0600
From:   Mike Christie <michael.christie@...cle.com>
To:     Wenchao Hao <haowenchao@...wei.com>, Lee Duncan <lduncan@...e.com>,
        Chris Leech <cleech@...hat.com>,
        "James E . J . Bottomley" <jejb@...ux.ibm.com>,
        "Martin K . Petersen" <martin.petersen@...cle.com>,
        open-iscsi@...glegroups.com, linux-scsi@...r.kernel.org
Cc:     linux-kernel@...r.kernel.org, liuzhiqiang26@...wei.com,
        linfeilong@...wei.com
Subject: Re: [PATCH] scsi:iscsi: Record session's startup mode in kernel

On 11/22/22 10:41 PM, Wenchao Hao wrote:
> Sorry I did not highlight the key points. The root reason we need to record
> node_startup mode in kernel is userspace's node_startup mode is unreliable in
> some scenarios:
> 
> 1. iscsi node and session is created in initrd, the configure files of these
>    nodes would be lost after we switch to rootfs
> 2. someone do iscsiadm -m discovery but did not specify the operation mode,
>    the iscsi node's node_startup would be updated to which specified in iscsid.conf
> 3. someone do iscsiadm -m node -o update to update nodes' configure
> 
> What's more, it seems "iscsiadm/iscsid" only refuse to logout of an ONBOOT
> session when logout is specificed by "--logoutall". We still can logout an
> ONBOOT session with "iscsiadm -m node -u comamnd".

logout_by_startup does go by the startup setting, but I think you missed the
session_in_use related code. It checks the mounts and holders already. Just
change it for whatever you need. I think your lvm use case should be covered
by the holder check. If not, add it.