linux-kernel - Re: [PATCH RFT] scsi: pm8001: Fix FW crash for maxcpus=1

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:   Thu, 6 Jan 2022 15:32:52 +0000
From:   John Garry <john.garry@...wei.com>
To:     <Ajish.Koshy@...rochip.com>, <damien.lemoal@...nsource.wdc.com>,
        <jejb@...ux.ibm.com>, <martin.petersen@...cle.com>,
        <jinpu.wang@...ud.ionos.com>, <Viswas.G@...rochip.com>
CC:     <linux-scsi@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <vishakhavc@...gle.com>, <ipylypiv@...gle.com>,
        <Ruksar.devadi@...rochip.com>,
        <Vasanthalakshmi.Tharmarajan@...rochip.com>
Subject: Re: [PATCH RFT] scsi: pm8001: Fix FW crash for maxcpus=1

On 06/01/2022 13:03, Ajish.Koshy@...rochip.com wrote:
>>   only a specific vector and, also, why we check at all in
>> an interrupt handler.
> Here is my initial understanding so far based on the code
> and data sheet
> 
> 1. Controller has the capability to communicate
> to the host about fatal error condition via configured
> interrupt vector MSI/MSI-X.
> 2. This capability is achieved by setting two fields
>   a. Enable Controller Fatal error notification
>      Dowrd 0x1C, Bit[0].
>      1 - Enable; 0 - Disable
>      Code: pm8001_ha->main_cfg_tbl.pm80xx_tbl.
>      fatal_err_interrupt = 0x01;
>   b. Fatal Error Interrupt Vector Dword 0x1C, bit[15:8]
>      This parameter configures which interrupt vector
>      is used to notify the host of the fatal error.
>      Code: /* Update Fatal error interrupt vector */
>      pm8001_ha->main_cfg_tbl.pm80xx_tbl.
>      fatal_err_interrupt |=
>      ((pm8001_ha->max_q_num - 1) << 8);
> 
> Probably this will be the reason why we check
> the vector in process_oq() for processing
> controller fatal error
> 
> if (vec == (pm8001_ha->max_q_num - 1)) {
>   
> Please do let me know if it helped in clarification.
> 

Sounds reasonable. And we only discover the issue for 8008/8009 now as 
we have that (pm8001_ha->max_q_num - 1) vector being used for standard IO.

So let me know of any other issue, otherwise I'll send a v2 with the 
coding style fixup.

Thanks,
John