linux-kernel - RE: [PATCH 3/4] net: dsa: microchip: fix interrupt mask

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]

Date:   Tue, 27 Aug 2019 20:56:54 +0000
From:   <Tristram.Ha@...rochip.com>
To:     <Razvan.Stefanescu@...rochip.com>, <andrew@...n.ch>
CC:     <Woojung.Huh@...rochip.com>, <UNGLinuxDriver@...rochip.com>,
        <vivien.didelot@...il.com>, <f.fainelli@...il.com>,
        <davem@...emloft.net>, <netdev@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH 3/4] net: dsa: microchip: fix interrupt mask

> On 27/08/2019 15:51, Andrew Lunn wrote:
> >
> > On Tue, Aug 27, 2019 at 12:31:09PM +0300, Razvan Stefanescu wrote:
> >> Global Interrupt Mask Register comprises of Lookup Engine (LUE)
> Interrupt
> >> Mask (bit 31) and GPIO Pin Output Trigger and Timestamp Unit Interrupt
> >> Mask (bit 29).
> >>
> >> This corrects LUE bit.
> >
> > Hi Razvan
> >
> > Is this a fix? Something that should be back ported to old kernels?
> 
> Hello,
> 
> During testing I did not observed any issues with the old value. So I am
> not sure how the switch is affected by the incorrect setting.
> 
> Maybe maintainers will be able to make a better assessment if this needs
> back-porting. And I will be happy to do it if it is necessary.
> 

I do not think the change has any effect as the interrupt handling is not implemented in the driver, unless I am mistaken and do not know about the new code.

Currently those 3 interrupts do not do anything that are required in normal operation.

The first one LUE_INT notifies the driver when there are learn/write fails in the MAC table.  This condition rarely happens unless the switch is going through stress test.  When this interrupt happens software cannot do anything to resolve the issue.  It may become a denial of service if the MAC table keeps triggering learn fail.

The second one is used by PTP code, which is not implemented.

The third one is triggered when register access space does not exist.  It is useful during development so driver knows it is accessing the wrong register.  It can also become a denial of service if someone keeps accessing wrong registers.  But then that person can do anything with the chip.